Mechanisms of Apoptotic Induction By Iron Chelators
Kirsteen Helen Maclean
A thesis submitted to the University of London For the degree of Doctor of Philosophy
1999
The Department of Haematology University College London Medical School University College London 98 Chenies Mews LONDON WC1E6HX ProQuest Number: U642343
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest.
ProQuest U642343
Published by ProQuest LLC(2015). Copyright of the Dissertation is held by the Author.
All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code. Microform Edition © ProQuest LLC.
ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 Abstract
The orally bioavailable hydroxypyridinone iron chelator, CP20 (LI or deferiprone) is known to induce bone marrow hypoplasia and thymic aplasia in laboratory animals and apoptosis in thymocytes and leukaemic cell lines, although the mechanisms are unclear. Experiments contained within this thesis have sought to elucidate how iron chelators CP20 and Desferrioxamine (DFO) induce apoptosis in murine thymocytes, human leukaemic cells and haemopoietic progenitor cells. Inhibition of the iron (III) containing enzyme, ribonucleotide reductase (RR) with consequent inhibition of DNA synthesis has been examined as a possible mechanism of apoptotic induction. The apoptotic effects of the RR inhibitor, hydroxyurea, have been compared with those of the iron chelators in thymocytes, human leukaemic HL60 cells and human haemopoietic progenitors. Apoptosis has been compared in different cell types by quantitative flow cytometry. DNA synthesis inhibition has been assessed by the incorporation of both BrdU and ^H-thymidine. Whereas iron chelators induce thymocyte apoptosis as early as 4 hours, hydroxyurea showed no effect, suggesting that RR inhibition is not the primary apoptotic mechanism in this cell type. By contrast, both the chelators and hydroxyurea induced apoptosis in proliferating HL60 cells where BrdU analysis showed that for both HU and chelators the apoptotic population was derived from cells which had recently entered S phase. In haemopoietic progenitors derived from CD34^ peripheral blood cells in liquid culture, apoptosis was induced by HU and chelators only when the cells were in cycle (> days 2 or < 9 days of culture). These findings are consistent with inhibition of RR being causative in apoptotic induction in proliferating cells but not in thymocytes. In thymocytes, induction of apoptosis by chelators requires RNA and protein synthesis because actinomycin D and cycloheximide respectively abrogate this process. Chelator induced apoptosis was equal in p53 knockout and wild-type thymocytes, suggesting that primary DNA damage is not the apoptotic trigger. A possible link between chelation of zinc and the induction of apoptosis was also investigated. In thymocytes, zinc was shown to abrogate the apoptotic effects of chelators in vitro and in vivo. Furthermore prolonged exposure of thymocytes to chelators deprives the cells of intracellular zinc, indicating that zinc chelation may contribute to the apoptosis. The bidentate hydroxypyridinones interact with intracellular zinc pools at low concentrations (l|iM CP20) in a fundamentally different manner from the hexadentate iron chelator DFO. Unlike the latter chelator, CP20 can shuttle zinc from inaccessible sites within cells onto larger zinc chelating molecules thereby enhancing apoptosis. In conclusion, the findings in this thesis show that proliferating cells in S- phase are particularly susceptible to apoptotic induction by iron chelators. Furthermore because of the similarity in terms of cell specificity and kinetics of apoptosis between HU and iron chelators, inhibition of RR is a likely mechanism of apoptotic induction in proliferating cells. However in thymocytes which are predominantly non-proliferating, a different mechanism of apoptotic induction must be invoked, which may in part involve the chelation of zinc. To Mum and Dad
For your constant love and support
“Zeus who leads mortals on the road to understanding, Zeus who has ordained that wisdom comes through sujfering.
Aeschylus-Agamemnon
If I should labor through daylight and dark Consecrate, valorous, serious, true. Then on the world I may blazon my mark; And what if I don't and what if I do ?
Philosophy by Dorothy Parker Acknowledgements
A number of people have contributed both directly and indirectly to the production of this thesis and I would like to take this opportunity to express my gratitude to them.
To.
Novartis Pharmaceuticals, Basle, Switzerland for providing me with the funding for this work
Dr. John Porter for his supervision, constructive critisism and giving me the opportunity to undertake such an interesting project
Professor David Linch for his support in reading the manuscript
Dr.s Rosemary Gale, Shaun Thomas and Asim Khwaja for reading the manuscript and helpful discussion
Dr. Pam Roberts for her 'tea and sympathy' chats
Mike Watts and Stuart Ings, thanks for the valuable CD34+ cells and all your computer know how
Arnold Pizzey for his flow cytometry and computer expertise but more importantly for keeping me in touch with reality.
Miss Louisa Hernandez for her calming words of wisdom
The rest of the Dept. Haematology (you know who you are!) thank you for sparing me a few minutes of your time on the way to the toilet to ask about my well being and for asking the golden questions, 'Need a drink?', or 'How about a fag?', I am indebted to you all.
Miss Katie McDonald for being my soul sister, a shoulder to cry on and keeping me sane during this frought period.
All my friends outside of work, but at the end of a telephone (in particular Natasha, Adam and Steve) thank you for providing me with your kind words of encouragement
Finally, my long-term love affair. Classic MAC, burglar proof, forever faithful??, the strong silent type. Will you ever forgive me for making you redundant now? INDEX PAGE
TITLE PAGE 1 ABSTRACT 2 DEDICATION 4 ACKNOWLEDGEMENTS 5 INDEX 6 LIST OF FIGURES 13 LIST OF TABLES 16 ABBREVIATIONS 17
CHAPTER 1 GENERAL INFORMATION 20
1.1 INTRODUCTION 21 1.2 IRON BIOLOGY 21 1.3 IRON HOMEOSTASIS 2 3 1.3.1 Transferrin and iron transport 24 1.3.2 Ferritin and iron storage 2 6 1.3.3 Iron-dependent transcriptional control of ferritin and transferrin 2 7 1.3.4 Iron absorption 29 1.3.4.1 Intestinal cellular iron absorption 29 1.3.4.2Non-intestinal cellular iron uptake 31 1.4 IRON OVERLOAD 33 1.4.1 Iron mediated radical damage 3 3 1.4.2 Causes of iron overload 3 5 1.4.2.1 Idiopathic haemochromatosis 3 6 1.4.2.2Thalassaemia 36 1.4.2.2.1 a-Thalassaemia 37 1.4.2.2.2 p-Thalassaemia 3 8 1.5 IRON CHELATION 3 9 1.5.1 Sites of iron chelation 4 0 1.5.1.1 Extracellular iron 40 1.5.1.2Intracellular iron 4 0 1.5.2 Chelation therapy 41 1.5.2.1 Desferrioxamine 43 1.5.2.2Hydroxypiridinones 46 1.5.2.3Clinical consequences of chelation therapy 48 1.6 APOPTOSIS 49 1.6.1 Morphological characteristics of apoptosis 50 1.6.2 Biochemical characteristics of apoptosis 50 1.6.3 Apoptosis vs Necrosis 52 1.6.4 Apoptosis and disease 54 1.7 MOLECULAR BIOLOGY OF APOPTOSIS 55 1.7.1 Role of Caenorhabditis Elegans 55 1.7.2 Role of proteases in apoptosis 57 1.7.3 Substrates cleaved by caspases during apoptosis 59 1.7.4 Inhibitors of caspases 60 1.7.5 A caspase hierarchy 62
CHAPTER 2 GENERAL MATERIALS AND METHODS 65
2.1 GENERAL MATERIALS 6 6 2.2 IRON CHELATORS 66 2.3 CELL CULTURE 67 2.3.1 Cell lines 67 2.3.2 Cell counting and viability 67 2.3.3 Freezing and thawing of cells 68 2.3.4 Cytospin preparation of cells 68 2.3.5 Thymocyte isolation 68 2.3.6 Murine neutrophil isolation 68 2.3.7 CD34^ isolation 69 2.4 MEASUREMENT OF APOPTOSIS 70 2.4.1 Flow cytometric analysis 70 2.4.2 Agarose gel electrophoresis 70 2.5 MEASUREMENT OF INTRACELLULAR ZINC 71 2.6 DETERMINATION OF HAEMOPOIETIC SUBSETS 73 2.7 STATISTICAL ANALYSIS 73 CHAPTER 3 APOPTOTIC INDUCTION IN VARIOUS CELL TYPES BY IRON CHELATORS 74
3.1 INTRODUCTION 75 3.2 APOPTOTIC INDUCTION BY IRON CHELATORS IN THYMOCYTES 76 3.2.1 Rationale 76 3.2.2 General experimental procedure 7 6 3.2.3 Demonstration of apoptotic features by light microscopy 7 7 3.2.4 DNA fragmentation by agarose gel electrophoresis 7 7 3.2.5 Quantitative analysis of DNA fragmentation by flow cytometry 7 9 3.2.6 Identification of apoptosis by immunoanalysis in CD4+CD8+ thymocytes 81 3.2.7 Time dependence of chelator-induced apoptosis 8 4 3.2.8 Concentration dependence of chelator-induced apoptosis 8 7 3.2.9 Discussion 87 3.3 APOPTOTIC INDUCTION BY IRON CHELATORS IN CD34+ CELLS IN CULTURE 91 3.3.1 Rationale 91 3.3.2 Isolation and culture of CD34 cells 9 6 3.3.3 Changes in CD34 cells under conditions of culture, proliferation, differentiation and cell cycle profile 9 6 3.3.3.1 Changes in morphology under conditions of culture 9 6 3.3.3.2Changes in cell number under conditions of culture 9 9 3.3.3.3Changes in surface antigen expression under conditions of culture 9 9 3.3.3.4Changes in cell cycle status under conditions of culture 102 3.3.4 Effect on apoptosis of continuous exposure to chelators as measured by quantitative flow cytometry 102 3.3.5 Effect of duration in culture prior to incubation with chelators on apoptosis after 24h exposure 104 3.3.6 Effect of chelator concentration on apoptosis following 24h exposure at various time points after initiation of CD34+ culture 104 3.3.7 Pulsed chase effect of 6h exposure to chelators: Effect on subsequent apoptosis 108 3.3.8 Discussion 111
8 3.4 GENERAL DISCUSSION 113
CHAPTER 4 THE RELATIONSHIP BETWEEN THE ANTI PROLIFERATIVE AND APOPTOTIC ACTIONS OF IRON CHELATORS 115
4.1 INTRODUCTION 116 4.2 EFFECT OF DNA SYNTHESIS INHIBITION ON THYMOCYTE APOPTOSIS 120 4.2.1 Rationale 120 4.2.2 Comparison of the effects of hydroxyurea and chelators in the induction of thymocyte apoptosis using flow cytometric analysis 121 4.2.3 Effect of iron chelators on thymocyte DNA synthesis using 3H thymidine uptake 124 4.3 EFFECT OF IRON CHELATORS ON PROLIFERATING LEUKAEMIC CELL LINES 126 4.3.1 Rationale 126 4.3.2 Effects of HU and chelators on apoptosis and DNA synthesis in HL60 cells 127 4.3.2.1 Effects of chelators and hydroxyurea on apoptosis and the cell cycle in HL60 cells using flow cytometric analysis 128 4.3.2.2Effect of time and concentration on apoptosis and the cell cycle of HL60 cells following chelator or hydroxyurea treatment 128 4.3.2.3 Analysis of the point in the cell cycle at which cells are committed to apoptosis by chelators and HU: BrdU analysis of chelator-treated HL60 cells 136 4.4 COMPARISON OF APOPTOTIC INDUCTION IN HAEMOPOIETIC CELLS BY HYDROXYUREA AND IRON CHELATORS 146 4.4.1 Rationale 146 4.4.2 Effect of hydroxyurea and iron chelators on apoptosis in committed murine granulocytes 147 4.4.3 Effect of hydroxyurea and iron chelators on apoptosis and cell cycle in CD34+ cells and their progeny in continuous culture 151 4.4.3. IRelative effect of continuous exposure of CD34^ cells to chelators and hydroxyurea on apoptosis 151 4.4.3.2Effect of continuous exposure of CD34^ cells to chelators and hydroxyurea on the cell cycle 153 4.4.4 Effect of transient exposure of CD34^ cells to hydroxyurea and chelators 153 4.4.5 Effect of chelator concentration on the cell cycle status of CD34^ cells 158 4.5 DISCUSSION 160
CHAPTER 5 INVOLVEMENT OF ZINC IN IRON CHELATOR-INDUCED THYMOCYTE APOPTOSIS 164
5.1 INTRODUCTION 165 5.2 EFFECT OF IRON CHELATOR ON INTRACELLULAR ZINC LEVELS 169 5.2.1 Rate of fall in intracellular zinc with addition of CP20 and DFO 170 5.2.2 Effect of pre-incubation of cells with iron chelators in subsequent zinquin fluorescence 173 5.3 EFFECT OF IN VITRO ADDITION OF CHELATOR-INDUCED APOPTOSIS 177 5.4 EFFECT OF DIETARY ZINC SUPPLEMENTATION ON CHELATOR- INDUCED APOPTOSIS IN VIVO 179 5.5 RELATIVE APOPTOTIC EFFECTS OF ZINC AND IRON CHELATORS 183 5.5.1 Effect of TPEN alone on thymocyte apoptosis and intracellular zinc levels 183 5.5.2 Effect of iron loading of thymocytes on apoptotic induction by zinc and iron chelators 184 5.5.3 Interactions of zinc and iron chelators 190 5.6 THE EFFECT OF IRON AND ZINC CHELATORS ON THE ACTIVITY OF THE ZINC CONTAINING ENZYME, PHOSPHOLIPASE C 194 5.7 GENERAL DISCUSSION 201
1C CHAPTER 6 GENETIC CONTROL OF CHELATOR-INDUCED APOPTOSIS IN THYMOCYTES 206
6.1 INTRODUCTION 207 6.2 ROLE OF RNA AND PROTEIN SYNTHESIS 209 6.2.1 Rationale 209 6.2.2 Experimental procedure 209 6.2.3 Results 210 6.2.4 Discussion 210 6.3 ROLE OF P53 IN CHELATOR-INDUCED THYMOCYTE APOPTOSIS 212 6.3.1 Rationale 212 6.3.2 P53 status 214 6.3.3 Cell type 214 6.3.4 Results and discussion 215 6.3.4. lEffect of p53 on thymocyte apoptosis 215 6.3.4.2Effect of p53 in murine granulocytes 215 6.4 ROLE OF NON-CASPASE PROTEASE INHIBITORS IN THYMOCYTE APOPTOSIS 218 6.4.1 Rationale 218 6.4.2 Experimental procedure 219 6.4.3 Results and discussion 219 6.4.3. lEffect of serine and aspartic protease inhibitors on iron-chelator induced thymocyte apoptosis 219 6.4.3.2Effect of the neutral protease inhibitor Calpain on chelator induced thymocyte apoptosis 221 6.5 EFFECT OF CYSTEINE PROTEASE (CASPASE) INHIBITORS ON APOPTOSIS 224 6.5.1 Rationale 224 6.5.2 Effect of caspase inhibitors on chelator induced apoptosis 225 6.6 GENERAL DISCUSSION 230
11 CHAPTER 7 DISCUSSION AND CONCLUSIONS 233
7.1 INTRODUCTION 234 7.2 INVESTIGATION OF CHELATOR INDUCED APOPTOSIS 234 7.3 MECHANISMS OF CHELATOR-INDUCED APOPTOSIS 236
CHAPTER 8 REFERENCES 239
12 INDEX OF FIGURES
Page CHAPTER 1 1.1 Iron homeostasis 24 1.2 Receptor-mediated endoeytosis 26 1.3 Translational control of transferrin and ferritin 28 1.4 Absorption of iron 30 1.5 Interacting pathways in iron uptake 32 1.6 Involvement of iron in cell injury and death 34 1.7 Worldwide prevalence of Thalassaemia 37 1.8 Treatments for Thalassaemia 39 1.9 Co-ordination of iron 42 1.10 Structure of desferrioxamine 44 1.11 Structure of CP20 44 1.12 Apoptosis vs Necrosis 53 1.13 CED genes in apoptosis 56 1.14 Activation of the caspase cascade 58 1.15 The caspase cascade 64
CHAPTER 2 2.1 Zinquin standard curve 72
CHAPTER 3 3.1 Morphology of thymocyte apoptosis 78 3.2 Agarose gel electrophoresis of chelator-induced apoptosis 80 3.3 Flow cytometric profiles 82 3.4 Effect of iron chelation on thymocyte apoptosis at 24h 83 3.5 Expression of CD4^CD8^ thymocytes in suspension culture 85 3.6 Effect of chelator concentration on thymocyte apoptosis at 24h 88
13 3.7 Haemopoiesis 92 3.8 Stem cell/progenitor ceil hierarchy 94 3.9 Morphology of CD34^ differentiation 97 3.10 Effect of differentiation on cell number of CD34^ cells 100 3.11 Effect of differentiation on antigen expression of CD34^ cells 101 3.12 Flow cytometric profiles of CD34^ cells 103 3.13 A) Apoptosis in CD34^ cells with continuous exposure to chelator 105 B) Effect of timing of exposure of CD34^ cells to chelation 106 3.14 Effect of pulse of chelator on apoptosis of CD34^ cells 109
CHAPTER 4 4.1 Flow cytometric profile of the cell cycle 116 4.2 Pathway of DNA synthesis 118 4.3 Model of ribonucleotide reductase 119 4.4 A) Comparison of chelators and HU on thymocyte apoptosis 122 B) Effect of hydroxyurea concentration on thymocyte apoptosis 123 4.5 Ih pulse of chelators/^H-thymidine measurement 125 4.6 Effect of chelators and hydroxyurea on HL60 cells 129 4.7 A) Effect of incubation time on apoptosis in HL60 cells 130 B) Effect of pulse on apoptosis in HL60 cells 133 C) Effect of concentration on apoptosis in HL60 cells 134 4.8 DNA-BrdU histogram 138 4.9 Kinetics of DNA-BrdU 141 4.10 Measurement of apoptosis in murine neutrophils 149 4.11 Effect of chelators and hydroxurea on murine neutrophils 150 4.12 A) Effect of continuous exposure of chelators and hydroxyurea on CD34+ cells 152 B) Effect of transient exposure of chelators and hydroxyurea on CD34+ cells 156
14 CHAPTER 5 5.1 Relationship of zinc to the cell cycle 168 5.2 Zinquin standard curve 171 5.3 Rate of fall of intracellular zinc with addition of chelators 172 5.4 Pre-loading of chelators on Zinquin fluorescence 175 5.5 Effect of zinc concentration 180 5.6 A) Effect of TPEN on thymocyte apoptosis 185 B) Effect of TPEN concentration on apoptosis 185 C) Effect of TPEN incubation time on apoptosis 186 D) Effect of TPEN on Zinquin fluoresence 186 5.7 A) Effect of addition of iron on chelator-induced apoptosis 188 B) Effect of addition of iron on zinquin fluoresence 189 5.8 Drug concentration combinations of TPEN/CP20 192 5.9 Spéciation plots of TPEN/CP20 195 5.10 Effect of Phospholipase C concentration on NPPC hydrolysis 197 5.11 Effect of chelators on Phospholipose C activity 198 5.12 Effect of chelator combinations on Phospholipase C activity 200
CHAPTER 6 6.1 Effector machinery involved in apoptosis 208 6.2 A) Effect of cycloheximide on chelator-induced apoptosis 211 B) Effect of Actinomycin D on chelator-induced apoptosis 211 6.3 Role of P53 in apoptosis 213 6.4 Effect of serine/aspartic protease inhibitors on apoptosis 220 6.5 Effect of calpain on chelator-induced apoptosis 222 6.6 Interacting pathways of apoptosis 223 6.7 Effect of ZVAD on chelator-induced apoptosis 227 6.8 A) Effect of AC-DEVD-CMK on chelator-induced apoptosis 228 B) Time course of AC-DEVD-CMK: Effect on apoptosis 228 6.9 Glucocorticoid-induced death 230
15 LIST OF TABLES Page
1.1 Iron-containing proteins and their function 22 1.2 Disorders associated with iron overload 35 1.3 Ideal properties of iron chelators 42 1.4 Apoptosis and disease 54 1.5 Caspases and their inhibitors 61
3.1 Effect of incubation time on chelator induced thymocyte apoptosis 86 3.2 Effect of chelator cone on CD34^ cell apoptosis 107
4.1 A) Effect of chelators and hydroxyurea on HL60 cell cycle 131 B) Effect of chelator cone on HL60 cell cycle 135 4.2 Effect of chelators and hydroxyurea on the cell cyele kinetics of HL60 cells: BrdU analysis 145
4.3 Effect of chelators and hydroxyurea on CD34^ (continuous exposure) 154
4.4 Effect of chelators and hydroxyurea on CD34^ (transient exposure) 157 4.5 Effect of chelators cone on eell cycle status of CD34^ cells 150
5.1 Importance of zinc 166 5.2 Concentration dependenee on the rate of fall of intraeellular zine 174 5.3 In vitro addition of zinc 178 5.4 In vitro addition of zinc 182 5.5 Combination of TPEN/CP20 191 5.6 Combination of TPEN/DFO 194 5.7 Effect of chelator coneentration on PLC activity 199
6.1 Effect of P53 status on chelators induced thymocyte apoptosis 216 6.2 Effect of P53 status on chelators indueed granulocyte apoptosis 217
16 LIST OF FREQUENTLY USED ABBREVIATIONS
ADP Adenosine diphosphate ATP Adenosine triphosphate BFU-E Burst forming unit-erthyroid BrdU 5’Bromo-2’ deoxyuridine BSA Bovive serum albumin CPU Colony forming unit CP20 1,2-dime thy 1-3-hydroxypyridin-4-one CPM Counts per minute dATP Deoxyadenosine triphosphate dCTP Deoxycytidine diphosphate DEX Dexamethasone DFO Desferrioxamine B dGTP Deoxyguanosine triphosphate DMSG Dimethyl sulphoxide DNA Deoxyribonucleic acid dNTP Deoxynucleoside triphosphate dTTP Deoxythymidine triphophate dUMP Deoxyuridine monophosphate EDTA Ethylene diamine tetraacetic acid EPR Electron spin (paramagnetic) resonance PCS Foetal calf serum GDP Guanosine diphosphate GI Gastrointestinal HbA Adult haemoglobin HbP Foetal haemoglobin Hb Haemoglobin HBSS Hanks buffered salt solution HC Hereditary haemochromatosis HCL Hydrochloric acid HLA Human leucocyte antigen HPO 3-Hydroxypyridin-4-one HU Hydroxyurea
17 IBE Iron binding equivalents
I C 5 0 Inhibitory concentration 50% ICE Interleukin 1 p converting enzyme IL3 Interleukin-3 IL6 Interleukin-6 IRE Iron regulatory element IRE-BP Iron regulatory element binding protein i.v Intraveneous LIP Labile iron pool LMW Low molecular weight MCV Mean cell volume Mwt Molecular weight NADH Nicotinamide adenine dinucleotide (reduced form) NADPH Nicotinamide adenine dinucleotide phosphate (reduced form) NTBI Non-transferrin bound iron P Partition coefficient PARP Poly(ADP-ribose)polymerase PBS Phosphate buffered saline PCD Programmed cell death Pen/Strep Penicillin/Streptococcus PI Propidium Iodide PKC Protein kinase C PLC Phospholipase C R Correlation coefficient RBC Red blood cell RE Reticuloendothelial RNA Ribonucleic acid RR Ribonuleotide reductase RT Room temperature s.c Subcutaneous SCF Stem cell factor SD Standard deviation SDS Sodium dodecyl sulphate TEMED N,N,N,N’ tetrethyleneethlenediamine Tf Transferrin TNF Tumour necrosis factor
18 TPEN N’N’N’N- Tetrakis(2-pyridylmethyl)ethylenediamine UDF Uridine diphosphate UTR Untranslated region
19 CHAPTER 1
General Information
20 1.1 Introduction
The aims of the thesis have been to study the potential mechanisms by which iron chelators induce apoptosis or programmed cell death in cell types such as murine thymocytes, primaiy human haemopoietic cells and various leukaemic cell lines. This follows the original observations that the orally absorbed hydroxypiridinone (HPO) iron chelator, CP20 is known to induce bone marrow hypoplasia and thymic involution in laboratory animals and agranulocytosis in thalassaemic patients, although the mechanism(s) are unclear. Iron overload is a major cause of morbidity and mortality in patients who receive regular blood transfusions. Currently, the predominant chelator in clinical use is desferrioxamine (DFO), but it has many disadvantages, including the need for sub cutaneous administration. An orally active iron chelator is therefore highly desirable and one such candidate is the hydroxypyridinone, CP20. However, CP20 has been shown to induce apoptosis in thymocytes and leukaemic cell lines. My studies have investigated the mechanisms by which this may occur. The introduction that follows provides an overview of iron biology and iron homeostasis coupled with a review of the advantages and disadvantages of iron chelation therapy.
1.2 Iron Biology
Iron (Fe) is the most abundant transition metal in living organisms and is essential to all forms of life. The range of reactions in which iron has an indispensable role spans all biochemistry. It is involved in the transport of oxygen by haemoglobin; in electron-transfer reactions, including the pathways of oxidative phosphorylation; in the synthesis of DNA (as an essential component of ribonucleotide reductase); in the catalysis of oxidation by oxygen and hydrogen peroxide; in the decomposition of noxious derivatives of oxygen, eg. peroxide and superoxide; and in many other reactions too numerous to mention. Iron is an essential component of many enzymes and oxygen binding proteins, which may be classified into 3 groups; haemoproteins, iron-sulphur proteins and non- haem non-iron-sulphur proteins and examples of proteins in each of these groups with their respective functions are shown in Table 1.1. Iron in aqueous solution has ready access to two oxidation states, the divalent ferrous form, Fe(II) and the trivalent ferric form, Fe(III) (Templeton, 1995; Ponka et ciL, 1998) which respectively can donate or accept electrons. However, unless appropriately coordinated, iron, due to its catalytic action in these one-electron redox
21 reactions, plays a key role in the formation of harmful oxygen radicals that ultimately cause peroxidative damage to cell structures (Meneghini, 1997). Both forms of iron form salts with common anions, but ferrous salts, although quite stable in the solid state, are oxidised to the ferric form in solution in the presence of oxygen, similarly ferric salts are stable as solids, however at neutral pH they form insoluble polynuclear hydroxide complexes which make the iron biologically unavailable. Several strategies must therefore be adopted for the acquisition of iron. Bacteria and fungi mainly acquire iron by the production of siderophores, small organic molecules with a very high affinity for Fe(III), which act as scavengers (Harrison et al., 1996). Higher animals eg. man acquire iron from foodstuffs. Plants tend to be a poor source of iron because of the prescence of phosphates, phytates and polyphenols, all of which inhibit absorption by formation of insoluble complexes. Meat is a better source of iron because the haem is readily absorbed (Harrison et at., 1996; Crichton et a l, 1998).
Protein/Enzyme Function
Haemoproteins
Haemoglobin Oxygen Transport
Myoglobin Oxygen Transport
Peroxidases Constitute microbicidal system of phagocytic cells
Catalases Decomposes H 2 O2 to O 2
Cytochrome C oxygenase Utilization of O 2
Cytochrome P450 Electron Transport Iron Sulphur Proteins
Ferredoxins Conversion of cholesterol to steroid hormones
NADH dehydrogenase Oxidative phosphorylation
Xanthine oxidase Electron transport
Aconitase Transforms citrate to isocitrate in the Kreb's cycle Non-Haem-Non-Suiphur Proteins
Ribonucleotide Reductase Reduces dTTP to TI P
Table 1.1 Examples of iron-containing proteins and their function
22 Because of this virtual insolubility and potential toxicity, iron is almost always found bound to specialized proteins either for transport (transferrin)(section 1.3.1), for storage (ferritin) (section 1.3.2) or as a component of functional compounds such as various enzymes, particularly for redox reactions (Brittenham et a l, 1994). This ability of iron to form coordination compounds is responsible for its binding to many organic molecules. In most of these, the iron atom forms an octahedral complex. One or more of the coordination sites may be linked to water molecules, or to molecular oxygen (which is important for the oxygen-transporting property of haemoglobin). Coordination may also occur to small molecules giving rise to iron chelates. These chelates, which normally involve Fe(III), may be of moderate affinity eg. citrate, or of high affinity eg. Desferrioxamine (DFO). Under normal physiological circumstances therefore the quantity of iron in the body is carefully controlled to avoid the accumulation of amounts that can exceed the capacity of the body for safe storage and produce the potentially toxic reactions.
1.3 Iron homeostasis in man
In man the average iron content is estimated at 4-5 g of which about 1.0-1.5g represents the iron reserve (Crichtonet al., 1996; Ponka et al., 1998). The amount of storage iron normally remains relatively constant throughout life, the stores of males being in general greater than those of females. The largest pool of iron is present as haem in haemoglobin, where it is involved in oxygen transport. Therefore de novo synthesis of haemoglobin during erythropoiesis in bone marrow is the most prominent route of iron utilization. Considering that the average life span of an erythrocyte is 3 to 4 months, at most 1% of the iron is recycled everyday. The second most abundant haem protein is myoglobin, which carries oxygen in muscle and this protein accounts for about 5% of total body iron. Figure 1.1 shows a schematic representation of iron physiology in man. Nutritional iron is absorbed by the intestines in small amounts compared with the total iron pool present in various tissues and proteins. Exchange between tissues occurs via serum, where iron circulates bound to transferrin. Most of the transferrin bound iron is used for the synthesis of haemoglobin by developing red cells. Senescent erythrocytes are phagocytosed by the cells of the monocyte- macrophage (reticulo-endothelial) system that liberate haemoglobin iron and release it back to plasma transferrin at a rate that normally matches the rate of iron transport for erythropoiesis. However at any given time, only about 0.1% of the total iron is in this recycling pathway. Hence, 99.9% of the total iron is either incorporated in tissue- specific expressed proteins, or stored as ferritin.
23 MUSCLE. OTHER MONOCYTE- PARENCHYMAL MACROPHAGE CELLS SYSTEM
CIRCULATING I I Functional RED BLOOD Iron CELLS
^Storage Iron
I I Transport — 'iron
ERYTHROID MARROW
HEPATOCYTES
Figure 1.1: Schematic representation of body iron supply and storage in man. Transferrin-Tf Adapted from Brittenham etal., 1994
1.3.1 Transferrin and iron transport Transferrin (Tf) is a plasma glycoprotein with a molecular mass of 80kDa and belongs to a family of related iron-binding proteins that includes lactoferrin and melanotransferrin (formerly known as tumour antigen p97). Schade and Caroline in 1946 were the first to prepare Tf from human plasma and showed that it had bacteriostatic activity, although the physiological role of Tf as a bacteriostatic agent is currently less understood. The Tf molecule consists of a single polypeptide chain constituted by two similar, but probably non-identical, iron-binding sites located in the N- and C- terminal halves of the molecule. Binding of Fe(III) to each of the N- and C- lobes of Tf is accompanied by the release of three protons and concomitant binding of one carbonate or bicarbonate ion, which causes the protein to be more acidic than in its iron-free state (Ponka et al., 1998). X-ray crystallographic studies of both human lactoferrin (Anderson et at., 1989) and rabbit serum Tf (Bailey et al., 1989) indicate that the Fe atom is coordinated by four amino acid ligands, the phenolate oxygens of two tyrosine residues, an imidazole nitrogen of a histidine residue and a carboxylate
24 oxygen of an aspartic acid residue. These amino acids occupy four of the six octahedral sites around each Fe atom, leaving two cis positions to be filled by water and/or the anion. The liver is the main source of Tf but it is also synthesized in small amounts in the brain (Espinosa de los Monteros et ai, 1994) and lymph nodes (Djehaet a i,1993). Transferrins are the iron binding proteins in body fluids and have two major functions. By binding iron, Tf directs the metal to cells that express transferrin receptors which provide the sole physiologic route of entry into the cell for Tf-bound iron. With the exception of mature erythrocytes, Tf receptors are probably expressed on all cells, with the highest expression being found on haemoglobin synthesizing cells, placenta, neoplastic tissue and rapidly dividing normal cells (Galbraithet al., 1980; Sutherland et al., 1981; Ponka et al., 1998), suggesting that the Tf receptor is expressed in coordination with cell proliferation. The receptor consists of a disulphide-linked transmembrane glycoprotein homodimer having a molecular weight of 180kDa and each subunit binds one molecule of Tf. The reported association constants of transferrin and the receptor vary greatly falling between 10~^ and 10"^ mol/L, but this may be due to a difference in receptor affinity between different cell types. Controlled delivery of iron to tissues is achieved by receptor-mediated endocytosis of a transferrin-transferrin receptor complex (Aisen, 1994; Ponka et a l, 1997; Richardson et al., 1997) (Figure 1.2). In the first step of receptor-mediated endocytosis, Tf attaches to specific Tf receptors on the cell surface by a physiochemical interaction, a process that is not dependent on temperature and energy. However, by a temperature and energy dependent process, the Tf-receptor complexes are then internalized by the cells and enclosed within endocytic vesicles. Iron is ultimately released from the Tf within the endocytic vesicles by a temperature and energy dependent process that involves endosomal acidification (Morgan, 1981). The mechanism of iron transport through the plasma membrane remains poorly understood, however one likely candidate appears to be the natural resistance- associated macrophage protein-2 (Nramp-2), although its precise cellular location in various cell types remains to be determined (Gunshin et a i, 1997). Fleming et al., (1997) suggested that mutation of this protein caused decreased iron uptake by erythroid cells of mice with microcytic anaemia and more recently in the Belgrade rat (Fleming et al., 1998). Once the iron has passed through the membrane it then enters a poorly characterised compartment known as the intracellular labile iron pool, the iron can then be ultimately used metabolically as described in section 1.2 or stored in ferritin.
25 Extracellular Intracellular
Clathrin-Coated Pit
Receptor-Mediated IRP1 and IRP2 Endocytosis Regulation of Intracellular Iron Proton Metabolism H + Pump
Iron Transport H *-A TPase?
Intracellular Labile Iron Pool [Fe(ll)/Fe{lll) ?]
Low Molecular Weight Fe Transferrin Complexes ? R eceptor % H aem and ----- High Molecular Weight Diferric ^ ^ / Non-Haem Fe . Intermediates ? Transferrin Containing — >■ ((/ / Proteins
Iron Uptake F e into Ferritin f Apo Storage Transferrin Ferritin
Figure 1.2: Iron uptake from transferrin via receptor-mediated endocytosis in mammalian cells
Adapted from Ponka et ciL, 1998
1.3.2 Ferritin and iron storage
Ferritin is a protein whose only clearly defined function is in the sequestration and storage of iron. Since the discovery of ferritin over 60 years ago by Laufberger in
1937, it has been found to have a widespread distribution, being present in virtually all cells, but particularly in the spleen, liver and bone marrow. The molecule consists of a central inorganic ferric oxyhydroxide core, which is surrounded by a protein shell, apoferritin (Thiel, 1987) which can accommodate up to 4,500 atoms of iron in its internal cavity. The protein shell itself has a molecular mass of 450kDa with 24 symmetrically related subunits of two types, a light subunit (L-subunit) of about 19kDa
26 and a heavy subunit (H-subunit) of about 21kDa (Crichton and Ward, 1992). Once within the protein shell, iron is stored in the ferric state as ferric-oxyhydroxide phosphate. Under normal conditions intracellular iron is stored as ferritin, however under situations of iron overload, cells contain another form, haemosiderin, which is probably a degradation product of ferritin (Ponka et a l, 1998). Haemosiderin is essentially localised in siderosomes which represent enlarged lysosomes. In iron- overloaded syndromes such as primary and secondary haemochromatosis the iron content of haemosiderin can increase by up to 100-fold whereas ferritin increases its iron storage capacity by only 5-10-fold (Selden et a l, 1980). The mechanisms surrounding the loading and unloading of iron in cellular ferritin are currently poorly defined. In vitro evidence indicates that ferrous iron, which is incorporated into the shell much more efficiently than ferric iron, is oxidised and deposited after its association with the inner surface of the subunits (Cowan et a l, 1993). Ferritin is constantly broken down within the lysosomes of cells and much of the iron released is reincorporated into new ferritin molecules. However Pippardet a l, (1986) showed that degradation of the ferritin protein may be an important mechanism for the release of iron and its availability to chelating agents. When ^^Fe-ferritin was administered intravenously into rats it was shown to be rapidly taken up by hepatocytes. The ^^Fe was readily available for chelation by DFO, however chelation of 59pe failed when the animals were pre-treated with chloroquine which inhibits lysosomal catabolism of ferritin.
1.3.3 Iron-dependent transcriptional control of ferritin and Tÿ fcceptor Cellular iron homeostasis is achieved through regulation of the cellular expression of the transferrin receptor and ferritin synthesis. Research conducted on non-erythroid cells cultured in vitro revealed that iron-dependent regulation of both transferrin and ferritin occurs at the post-transcriptional level, mediated by iron- responsive elements (IREs). IREs were first identified in the 5' untranslated regions (UTRs) of ferritin H- and L- chain mRNAs (Klausner et a l, 1993; Theil, 1994; Hentze et a l, 1996) and were shown to mediate inhibition of ferritin mRNA translation in iron- deprived cells. The IREs are cw-acting nucleotide sequences, forming stem loop structures that contain an unpaired cytidine 6 bases 5' of a 6 membered loop whose sequence is CAGUGN. These hairpin structures are recognised by rra«5-acting cytosolic RNA binding proteins known as iron regulatory proteins (IRPs). The IRPs coordinate the intracellular uptake and storage of iron by translational control of the synthesis of transferrin receptor and ferritin and the interactions of the IRPs with IREs is well documented (Iwai et a l, 1995; Rouault and Klausner, 1996; Hentze et a l, 1996; Richardson et a l, 1997) (Figure 1.3).
27 TRANSFERRIN RECEPTOR mRNA FERRITIN mRNA LESS STABLE WELL TRANSLATED
5V AAAA
AAAA LOW AFFINITY IRP1JRP2
DECREASED IRON POOL ACTIVATES i INACTIVATES IRP * IRP INCREASED IRON POOL
HIGH AFFINITY IRP1JRP2 5' 5 AAAA 7 1 ^ AAAA
MORE STABLE NOT WELLTRANSLATED
Figure 1.3: Regulation of Transferrin receptor and Ferritin synthesis by IRPl and IRP2
Transferrin receptor synthesis is controlled by adjusting the amounts of cytoplasmic transferrin receptor mRNA. The 3' untranslated region (3'UTR) of transferrin receptor contains 5 IREs. Binding of IRPS tothe IREs in the 3'UTR retards cytoplasmic degradation, increasing the concentration of cytoplasmic transferrin receptor mRNA and the rate of transferrin receptor synthesis. With an increased number of cellular transferrin receptors, iron uptake is enhanced. By contrast, ferritin synthesis is controlled without changes in the amount of ferritin mRNA present by repressing translation of ferritin mRNA. The 5' untranslated region (5'UTR) of ferritin and mRNA contains a single IRE. Binding of an IRP to the IRE in the 5' UTR arrests translation of ferritin mRNA so less ferritin is produced and iron sequestration is diminished.
28 Recently two closely related IRPs which have been designated IRP-1 and IRP- 2, have been purified and cloned (Hentze et al, 1996; Richardson et al, 1997). Interestingly, IRP-1 was found to share homology with mitochondrial aconitase, an iron sulphur [4Fe-4S] cluster-containing enzyme of the Kreb's (citric acid) cycle. In iron-replete cells, IRP-1 contains a [4Fe-4S] cluster and in this form possesses aconitase activity and binds RNA with low affinity. In contrast when iron is scarce, IRP-1 lacks a [4Fe-4S] cluster and aconitase activity and binds to IREs with high affinity. IRP-2 lacks this aconitase activity and instead functions solely as an RNA- binding protein, the regulation of IRP-2 by iron being mediated by specific proteolysis. Transferrin receptor synthesis is regulated by controlling the stability of cytoplasmic transferrin receptor mRNA. The 3' untranslated region (3'UTR) of transferrin receptor mRNA contains 5 IREs. Binding of the IRP's to the IRE s in the 3 ' UTR retards cytoplasmic degradation, increasing the concentration of cytoplasmic transferrin receptor mRNA and hence transferrin receptor synthesis. With an increased number of cellular transferrin receptors, iron uptake is enhanced. By contrast, ferritin synthesis is controlled without changes in the amount of ferritin mRNA present by repressing translation of ferritin mRNA. The 5' untranslated regions (5'UTR) of ferritin and ferritin mRNA contains a single IRE. Binding of an IRP to the IRE in the 5'UTR arrests translation of ferritin mRNA resulting in less ferritin being produced and therefore iron sequestration is diminished. These mechanisms allow cells to maintain a tight control over intracellular iron pools.
1.3.4 Iron absorption For normal body homeostasis iron absorption must balance iron excretion. In man iron is absorbed from the gastrointestinal tract, primarily by the enterocytes within the duodenum. Recently new insights into the mechanism and regulation of intestinal iron absorption suggest the existence of an "Alternative Pathway" for iron uptake. Unlike the previously elucidated transferrin-transferrin receptor pathway, the alternative pathway does not involve the major iron-carrying plasma protein transferrin (Umbreitet a l, 1998) (Figure 1.5).
1.3.4.1 Intestinal cellular iron uptake Iron is predominantly found in the diet as haem iron and various iron salts eg. ferric citrate or as complexes with protein and carbohydrates (Conrad, 1987). In man normal individuals absorb between 1 and 2 mg each day from the duodenum and upper jejunum. Most dietary inorganic iron is ferric iron as iron in the ferrous valence state Fe(II) is spontaneously oxidised to the ferric state(III). In the acid pH of the stomach the iron is solubilised and it chelates mucins and certain dietary constituents to keep it soluble and available for absorption in the more alkaline duodenum (Conrad et al.
29 1993). Intestinal iron absorption can be divided into three distinct phases: luminal; mucosal and systemic. The luminal phase concerns the delivery of iron to the brush borders of the intestinal mucosal cells and is detemiined by the type of food ingested and by its interaction with intestinal secretions (Turnbull, 1974; Templeton, 1995; Crichton and Ward, 1996). Iron is found in the form of Fe (III) chelate in the lumen of the duodenum. On the mucosal surface the iron is tranferred to mucin (Conrad et ai, 1993) which acts as a solubilizing chelator (Figure 1.4). Mucosal uptake of iron is facilitated by a paraferritin complex [Paraferritin (PF), so called because it was found on chromatographic sizing columns in close proximity to ferritin (Umbreit et al., 1996)]. The complex contains the duodenal surface (33 integrin, flavin monooxygenase. The complex acts as a niccbnamide adenine dinucleotide phosphate-dependent femreductase and converts iron to the reduced redox state required for the use of iron in the synthesis of haem proteins.
ABSORPTION OF INORGANIC AND HAEM IRON
Fe (III) GUT Haemoglobin & Myoglobin (pH 2) LUMEN ^ Proteolysis Ascorbate Mucin Histidine Haem Fructose Bile salts ^ Amino Acids t Soluble haem chelate Soluble Fe|(III) chelate ABSORPTIVE Haem vesicle ? MEMBRANE (33 Integrin
Haem Fe (III) Mobilferrin Oxygenase
CYTOSOL + Paraferritin Paraferritin t I NADPH 1 NADPH Fe (II) Fe (II)
Figure 1.4: Absoiption of inorganic iron and haem. Although there are major differences in the absorptive pathways of inorganic iron and haem-iron, ultimately iron binds paraferritin and is reduced so it is available to the cell in the appropriate redox for use. Adapted from Uzel et a!., 1998.
30 It is currently unclear how haem traverses the membranes of absorptive surfaces of intestinal epithelial cells and targeted to specific intracellular locations for degradation of the haem with the release of inorganic iron (Figure 1.4) The absorptive mechanism for haem in intestinal micovilli is probubly different from cellular uptake of haem and/or haemoglobin by phagocytic cells. In plasma, haem and haemoglobin are bound to proteins (haptoglobin, albumin) and it is these haem and haemoglobin protein complexes that are transported into phagocytes to conserve body iron (Smith et al., 1981; Uzel et al, 1998). Unbound free haem and haemoglobin are excreted by the kidney and can produce iron deficiency if there is considerable loss of iron in the urine. These binding proteins do not exist in the gut lumen making it probable that the intestinal haem receptor is different than that which exists on phagocytic cells in the liver and other body organs. Recently, Umbreit et al, (1998) showed preliminary data postulating that the membranes of absorptive cells possess both halo-transferrin and apotransferrin receptors that regulate the ingress and egress of cellular iron. Not all the iron taken up from the lumen into the cells is transferred into the plasma, a variable proportion can be sequestered within mucosal cells and eventually discarded into the gastrointestinal tract when the cell exfoliates, and iron absorbed from the gut in excess of body requirements might be incorporated into mucosal cell ferritin (Hartman et al, 1963; Crichton, 1991). The systemic phase is concerned with the mechanisms which determine how the body's iron requirements are communicated to the mucosal cell. The uptake of iron from the lumen of the intestines increases when larger doses of iron are ingested (Charlton et al, 1965; Skikne et al, 1994). However it has been observed that with increasing concentrations of iron, the percentage of iron uptake decreases. Such a block, originally described by Hahn in 1943, has been referred to as a "mucosal block”. The mucosal block theory has been invoked to explain the control of iron uptake. The amount of iron in the enterocyte increased with iron repletion and decreased with iron deficiency suggesting that iron saturation of critical sites in the mucosal tissues controlled the uptake of iron (Ponka et al, 1998).
1.3.4.2 Non-intestinal cellular iron uptake Unlike absorptive cells in the intestine, non-intestinal cells appear to possess three pathways for uptake of inorganic iron, these being the classical transferrin- transferrin receptor pathway, the transferrin associated transferrin receptor independent pathway (TRIP), and the transferrin independent mobilferrin-integrin paraferritin pathway (MIP) (Figure 1.5). The TRIP is used when transferrin receptors become saturated at physiological concentrations of iron and Tf, again internalized mobilferrin acts as an intermediate between the Tf iron and haem (section 1.3.4.1). The recently
31 described alternative MIP pathway is the major pathway for intestinal absorption of inorganic iron and may only be used efficiently for mucosal uptake of iron and iron- overloaded individuals with fully saturated transferrin (Umbreit et al, 1998). Alternatively it may facilitate iron uptake from the TRIP after degradation of transferrin near the surface of the cell. Therefore, iron levels in normal individuals are under stringent control. However, there are certain conditions when the iron status can change, either locally as found in ischaemic tissue or systemically as with haemochromatosis or transfusion induced iron overload, where abnormal levels of iron can induce toxic symptoms.
Tf-Fe
Fe-chelates Tf-receptor Haem-Fe
Tf-Fe Haemoxygenase
Mb Proteolysis PF
Fragments ,2+
Tf-Fe Haem-Fe
Figure 1.5: Interacting pathways involved in iron uptake
Iren can be absorbed via four different pathways. These include two pathways in which iron is presented to the cell bound to transferrin: (A) the transferrin (Tf)- transferrin receptor clathrin-coated pit pathway and (B) the transferrin-receptor-independent pathway (TRIP). The transferrin independent pathways are,
(C) the mobilferrin-integrin pathway (MIP) and (D) the haem pathway.
Mb-Mobilferrin, PF-paraferritin.
Adapted from Ponka et ai, 1998
32 1.4 Iron overload
Whilst disturbances of iron balance usually result in a reduction of total body iron, in some situations iron overload occurs (Crichton, 1991). Iron overload develops in any circumstance in which there is a prolonged positive iron balance. Most of the surplus iron is deposited in the reticuloendothelial system and/or parenchymal cells. Evolution has given us mechanisms for absorbing dietary iron with about 10% efficiency, but we have no mechanism for the elimination of excess iron. If the body experiences a sudden and significant loss of blood, stores of iron from ferritin can be drawn on for the synthesis of new haemoglobin to replace that which is lost, usually however transfusion supplants this need. Therefore in a healthy individual, excessive iron stores serve no known function (McCord, 1998). Because humans have no physical means of eliminating excess iron, any sustained increase in the amount of iron entering the body may eventually result in iron overload. As suggested earlier (section
1 . 2 ), iron is usually made unavailable to participate in the generation of harmful free radicals by binding to ligands such as transferrin, however in the scenario of iron overload such protective mechanisms become saturated.
1.4.1 Iron mediated free radical damage Like most transition metals, iron in excess is a toxic burden to the organism. Although the mechanism by which iron causes tissue damage is essentially unknown, it seems likely that free radicals are involved. A free radical has been defined as any species capable of independent existence that contains one or more unpaired electrons. Indeed the reduction of molecular oxygen, 02 to water, where the products are superoxide (02"), hydrogen peroxide (H2 O2 ) and the hydroxyl radical (HO*) is one of the commonest reactions in the body. Many investigators believe that the most destructive action of the superoxide radical is in bringing about the reductive release of iron from ferritin (Biemond et al 1986, McCord, 1998). It has been proposed that O2. enters the ferritin core through the hydrophilic channels, followed by the reduction of Fe(III) to Fe(II). This enables release of iron from the ferritin core (Biemond et al, 1988). Once iron has been liberated in the presence of superoxide and its dismutation product, hydrogen peroxide, the hydroxyl radical (HO.) may be formed by Haber- Weiss chemistry:
Fe2+ + H2 O 2 » » > Fe3+ + OH + HO. 0 2 " + Fe3+ » » > O2 + Fe^ +
O 2 " ^ 2 ^ 2 O2 "I" OH“ + HO.
33 Unlike the superoxide radical, which is not highly reactive compared to most other free radicals, the hydroxyl radical is an extremely powerful oxidising species. The hydroxyl radical can attack all classes of biological macromolecules. It can depolymerise polysaccharides (McCord et al, 1974), cause DNA strand breaks (Halliwell et al, 1992), inactivate enzymes (Zhang et al, 1990) and initiate lipid peroxidation (Gutteridge et al, 1982) (Figure 1.6).
Disease Processes
Superoxi^ _ Glutathione Peroxidase Dismutase — Catalase Ferritin and ,2+ Fe/S proteins HO-
Lipid Peroxidation
Phospholipid hydrogero^de_l _ “ Vitamin E Vitamin C Glutathione Peroxidase X
Cell Injury and Death
Figure 1.6: Involvement of iron in cell injury and death The protective antioxidant enzymes and vitamins are shown in italics and their points of intervention in the process are indicated by the dotted lines. Adapted from McCord, 1998
Lipid peroxidation is a chain reaction that is amplified by redox-active iron and it is the action of the hydroxyl radical which would appear to have the greatest pathophysiological consequences in diseases such as ischaemic heart disease and
34 stroke. Furthermore, the increased susceptibility of membrane lipids to peroxidation as observed in thalassaemic red cells may be a function of the excessive iron seen in these patients. Free radicals have been implicated in iron-induced injury to many tissues, including lung (Adamson et a l, 1993) and heart (Scott et al, 1985; Hershko et al, 1987; Link et a l, 1993). Other tissues from iron-overloaded animals are also hypersensitive to peroxidative injury, again by a suspected free radical reaction. Indeed, iron-catalysed lipid peroxidation and its suppression by iron chelators have been used as a means of evaluating the potential usefulness of a chelator (Hershkoet al, 1987).
1.4.2 Causes of iron overload In both adults and children, iron overload may be produced by an increased absorption of dietary iron, by parenteral administration of iron or both. Increased absorption of iron may be the result of 1. an inappropriately elevated absorption of iron from a diet containing normal amounts of iron (as occurs in idiopathic haemochromatosis and iron-loading anaemias). 2. Consumption of large amounts of dietary iron, eg. In the South African Bantu population.(Gordeuket al, 1992). Large amounts of parenteral iron loading is produced by repeated blood transfusions, of which the thalassaemias represent the classic example. Disorders associated with iron overload are outlined in Table 1.2
Table 1.2: Disorders associated with iron overload
D isorder Mechanisms of iron overload
Hereditary haemochromatosis Recessive inheritance of increased intestinal iron absorption Iron-loadinig anaemias Ineffective erythropoiesis, with or without red cell transfusion Excessive iron intake Red cell transfusions Infusion o f haemoglobin iron Elemental iron Prolonged ingestion of medicinal iron African iron overload Ingestion of excessive dietary iron + genetic factor (?) Porphyria cutanea tarda Associated inheritance of hereditary haemochromatosis allele(s) Liver disease Increased intestinal absorption of dietary iron Focal haemosiderosis Idiopathic pulmonary Alveolar extravasation of red cells Renal Haemosiderosis Chronic intravascular haemolysis Iron overload in animals Experimental iron overload Infusion/feeding of excess elemental iron Experimental copper deficiency Ceruloplammin deficiency (impaired iron oxidation) p2 microglobulin deficiency Increased intestinal iron absorption (HFE gene defect)
Adapted from Bottomley, 1998
The two genetic disorders commonly associated with parenchymal iron overload which have been particularly well documented in the literature are the primary iron overload of idiopathic haemochromatosis, and the secondary iron overload of the
35 thalassaemias. The causes are respectively, excessive iron absorption from the diet, and parenteral administration of iron in the form of transfused red blood cells.
1.4.2.1 Idiopathic Haemochromatosis The term haemochromatosis was introduced over one hundred years ago by Von Recklinghausen to describe the combination of extensive haemosiderin deposits in the liver, accompanied by cirrhosis. The genetic basis of haemochromatosis was established when it was shown that that there was a close linkage between haemochromatosis and the human leucocyte antigen (HLA) class 1 loci, with the disorder being inherited as an autosomal trait (Simon et al, 1977). The phenotype may manifest at one extreme by organ damage due to iron overload (this phenotype would include cirrhosis, diabetes, arthropathy, endocrine failure, myocardopathy and skin pigmentation). At the other end of the phenotypic scale is the finding of increased transferrin saturation concentration with no evidence of iron overload or organ damage. Haemochromatosis is one of the most common inherited diseases in Caucasians of European descent. 10-13% of this population are heterozygotes for the haemochromatosis mutation with homozygotes occurring with a frequency of 5 in 1000 (Bothwell et al, 1998). Recognition of a conserved haplotype on haemochromatosis chromosomes led several groups to a positional cloning strategy in order to isolate the haemochromatosis gene (Jazwinska et al, 1995; Feder et al, 1996; Ajioka et a l, 1997). The gene subsequently has been designated HFE. The HFE gene encodes a HLA class 1-like molecule and a mutation in HFE leading to the substitution of tyrosine for cysteine at position 282 (C282Y) has been identified in most haemochromatosis homozygotes. The C282Y mutation occurs in a highly conserved region of HLA class 1 molecules, a region which is involved in p2-microglobulin binding. Recently, Rothenberg et al, 1996 described an association in P2 microglobulin and iron overload in p2 microglobulin knockout mice. It was found that the mice developed iron overload with ageing and the distribution of body iron stores was identical to that found in haemochromatosis in humans. The mechanism by which the C282Y mutation in HFE leads to iron overload remains to be established.
1.4.2.2 Thalassaemia In 1925 Thomas Cooley described a series of cases of splenomegaly in immigrant children with anaemia and peculiar bone changes. This was the first description of thalassaemia, a name first coined by Whipple and Bradford in 1936. The name is derived from the Greek word thalassa, meaning the sea, since the condition is particularly common to people originating from the Mediterranean basin. However the distribution of the thalassaemias is now known to be widespread, with an incidence
36 extending through Southern Europe and North Africa and the Middle East to India, Indonesia and the Far East (Figure 1.7). The thalassaemia syndromes represent the most common hereditary anaemias worldwide that are characterised by ineffective erythropoiesis and associated iron overload. The thalassaemias are associated with an impaired production of one or more of the normal polypeptide chains of globin. Any of the four definitive polypeptides that occur in normal haemoglobins may be involved (a, (3, y, ô). However the most prevalent thalassaemia syndromes are those involving the a or p globin chains designated a and p thalassaemia respectively (Weatherall and Clegg, 1981; Stumpf and Townsend, 1992).
|Ncw incas[~~~[ Otiyiiiol iircos
Figure 1.7: Worldwide prevalence of Thalassaemia showing areas of thalassaemia incidence Adapted from Dobbin et al, 1990
1.4.2.2.1 a Thalassaemia Normal people have twoa globin genes on each of their chromosome 16 (Crichton, 1991; Pippard, 1994). If both are lost (a- thalassaemia) no a globin chains are made, whereas if only one of the pair is lost (a+ thalassaemia) the output of a
37 globin chains is reduced. Impaired a globin production leads to excess y or (3 chains that form unstable and physiologically useless tetramers, 7 4 (Hb Bart's) and p4 (HbH). The homozygote state for a- results in the Hb Barts fetalis hydrops syndrome, whereas the inheritance of a- and a+ thalassaemia produces HbH disease. The Hb Barts hydrops fetalis syndrome is characterised by the stillbirth of a severely oedematous foetus in the second half of pregnancy. Hb H disease is associated with a moderately severe haemolytic anaemia (Beris et al, 1995).
1.4.2.2.2 p Thalassaemia The p thalassaemias result from over 150 different mutations of the P globin genes and occur in one in four offspring, if both parents are carriers of P thalassaemia (autosomal recessive). The p globin gene is located on chromosome 11. Heterozygotes for P thalassaemia are asymptomatic and have hypochromic microcytic red cells with a low mean cell haemoglobin. Homozygotes usually develop severe anaemia in the first year of life. This results from the deficiency of p globin chains causing excess a chains to precipitate in red cell precursors leading to their damage either in the bone marrow or peripheral blood. Hypertrophy of the ineffective bone marrow leads to skeletal changes. If patients are transfused, growth and development may be 'normal'. However, they accumulate iron and may die from damage to the myocardium, pancreas or liver. They are also prone to infection and folic acid deficiency (Weatherall, 1996). Currently the only effective treatment for thalassaemia is to increase the haemoglobin levels by frequent blood transfusions, without which most patients die within the first year of life. The goals of transfusion therapy are therefore to suppress excessive ineffective erythropoiesis, eliminate the complications of anaemia and compensatory bone marrow expansion, permit normal development throughout childhood and extend survival (Weatherall and Clegg, 1981). However one unit of blood corresponds to 250mg of iron which cannot be eliminated from the body. The progressive accumulation of iron from regular blood transfusions leads to cardiac, endocrine and liver damage, with death usually by the age of 2 0 years from heart failure or arrhythmias (Modell, 1979). To prevent or treat iron overload, chelation therapy is necessary (Figure 1.8).
38 Percentage survive Transfusion and Desfera I HOO 80 60 Transfusion only 40 No treatment -20
— 0 Age/years 0 10 20 30
F ig u re 1.8:Success of various treatments for thalassaemia Adapted from Dobbin et al, 1990
1.5 Iron chelation
In normal subjects, most of the body iron is unavailable for chelation and even in iron-overloaded individuals there are protective mechanisms that minimise the damage potentially arising from the increased iron load. As previously mentioned iron is stored as the relatively inert haemosiderin precipitate, ferritin is induced and the transferrin receptor is down-regulated to minimize iron transferrin-mediated uptake (section 1.3). On the other hand, uptake of non-transferrin bound iron is increased in iron-loaded cells (Parkes et al, 1993). Therefore, 39 1.5.1 Sites of iron chelation 1.5.1.1 Extracellular iron Iron can be chelated either in the plasma or from within cells. In the plasma, iron is to be found tightly bound to transferrin and as such is virtually unavailable for chelation in vivo. However in situations of increasing iron overload, the transferrin iron binding capacity becomes saturated and non-transferrin bound plasma iron (NTBI) appears. Although NTBI is quantitatively small (2.7-7. Ijimol/l in thalassaemic sera) it does represent an important source of iron, particularly in the generation of free radicals. However, as this iron is not coordinated it is easily chelatable (reviewed by Porter gr a/., 1989). The turnover of transferrin bound iron is of the order of 30mg/d and iron release from reticuloendothelial cells is therefore a major source of chelatable iron. Evidence for this comes from studies on the action of DFO in laboratory animals and in chnical studies. Using cell labelling techniques in iron overloaded rats and ferrokinetic data in humans several investigators have suggested that most urinary iron excretion with DFO is derived from this source (Hershko, 1975; 1978, Hershko et al., 1979). However others suggest that a proportion of the urinary iron may be derived from hepatocytes and other parenchymal cells (Pippardet al, 1982; Hershko et at., 1994). Theoretically an iron chelator may also compete with transferrin at the site of delivery of iron to the red cell precursors or hepatocytes (Porteret al, 1989) (Figure 1.2 ). 1.5.1.2 Intracellular iron Within the cell, iron chelators can potentially chelate from 3 kinds of iron pool: 1. storage iron (in ferritin and haemosiderin), 2. the LIP and 3. non-haem Fe-containing proteins (Figure 1.2). The main source of intracellular iron is released from ferritin and in the scenario of iron overload, would theoretically represent a potentially accessible source of chelatable iron for negative iron balance to be achieved. However, because of the way in which iron is stored within the ferritin core it is relatively difficult to chelate directly and indeed the slow rate at which ferritin iron can be chelated, especially by large chelators, limit its importance (Porter et al, 1989). Another important pool of chelatable iron is the transient, low molecular weight, so called LIP. The existence of such a pool has been demonstrated in many cell types including erythroid cells, Chang cells, peripheral blood leucocytes, reticuloendothelial cells, intestinal mucosa and hepatocytes. Currently, very little is known about the nature and properties of the low molecular weight intracellular iron pool (Porteret al., 1989; Templetonet al, 1995), although Cabantchik et al, (1996) and Epsztejn et al., (1997) have attempted to determine LIP levels in living cells based on the fluorescent 40 probe, Calcein (CA). Finally, there are functional iron pools bound to iron containing molecules which are essential for normal cellular function and include haemoglobin, myoglobin and enzymes such as lipoxygenase and ribonucleotide reductase. However, the iron contained within these molecules is not chelatable and iron levels are essentially the same in iron-overloaded as in normal subjects amounting to 3g (Porter et al, 1989). Therefore it would seem clear that it is the low molecular weight pool which represents the major compartment through which all intracellular traffic of iron passes and is readily chelatable. 1.5.2 Chelation therapy An iron chelator is an organic molecule containing two or more functional ligand groups which are capable of forming coordinate bonds with iron to form a heterocyclic iron-containing ring (a chelate). The shared electrons of the co-ordinate bonds are usually donated by atoms of nitrogen, oxygen or sulphur in the ligand groups. Fe(II) is regularly oxidised to virtually insoluble Fe(III) under physiological conditions. Fe(II) has a coordination of 6 therefore it is most stable when linked with six oxygen atoms. These may be supplied by 3 bidentate chelators donating two oxygen atoms each (Figure 1.9), hence bidentate chelators such as CP20 form a 3:1 complex with iron or the oxygens may be supplied by one hexadentate chelator donating all 6 oxygen atoms, hence hexadentate chelators such as DFO form a 1:1 complex with iron (Pippard 1989). The stability of a metal chelate in solution is influenced by the number of donor groups present on the same molecule according to the so-called chelate effect, ie. hexadentate ligands such as desferrioxamine form more stable complexes than the corresponding bidentate or tridentate ligands. Hexadentate ligands also have greater scavenging ability at low concentrations (<20|LiM) and are less likely to dissociate (Hider and Singh, 1992). The affinities between metal ion and ligand are usually expressed as the equilibrium constants for the formation of a complex from hydrated metal to the fully dissociated form of the ligand. Fe3+ + L < >FeL3+ K= [FeL3+]/[Fe3+] [L] The ability of a given chelator to remove iron from iron proteins such as transferrin is not only based on thermodynamic but also on kinetic considerations. Indeed, rates of removal of iron from transferrin may vary dramatically according to the chelator (Kretchmar Nguyen et al, 1993). Finally, iron mobilisation from iron containing proteins is also dependent upon tissue distribution, especially the membrane permeability of the chelator. The partition coefficient of the chelator between octanol and water at pH 7.4, measured as log p, gives an estimation of the facility to penetrate the cell membrane by simple diffusion as well as of potential oral activity (reviewed by 41 Galey, 1997). Table 1.3 outlines the factors which influence iron chelation. Simplistically an iron chelator should itself be non-toxic, specific for iron, suppress Fenton chemistry, remove iron from cellular pools and lead to its rapid excretion without redistribution. The goal of chelation therapy is to normalize the body iron load as rapidly as possible without toxicity and then to maintain this status. For the past 20 years, Desferrioxamine B has been a clinically useful drug available to relieve iron overload BIDENTATE TRIDENTATE HEXA DENTATE Figure 1.9: Coordination of iron Adapted from Hider et al, 1996 A. General properties 1. High affinity/selectivity for Fe(III) 2. Low affinity for other transition metals, eg Cu(II) 3. Oral effectiveness 4. Access to available iron pools 5. Long-half life of the free chelate 6 . Inexpensive 7. Long term tolerability B. Properties to minimise toxicity 1. Limited lipophilicity of the iron-free chelator 2. Rapid elimination of the iron ligand complex 3. No re-distribution/re-absorption of iron 4. No facilitation by iron-ligand complex of microbial growth Table 1.3: Ideal properties of an iron chelator 42 1.5.2.1 Desferrioxamine Desferrioxamine (Desferal, DFO) is a hydroxamate fungal siderophore produced by Streptomyces pilosus with a high affinity for iron (III)(log P=31) (Figure 1.10) By virtue of its high water solubility, the molecule can selectively scavenge iron and facilitate its excretion in urine and bile. DFO in vitro enters a variety of cells where it binds iron in the LIP which is in equilibrium with the storage pool (Fosburg, 1990). Ferrioxamine, the DFO-iron complex, exits the cells with varying efficiencies depending on the cell type. DFO in vivo is largely confined to the extracellular space. However, it seems to enter hepatocytes quite readily and the ferrioxamine that is formed appears almost exclusively in the stool, suggesting an unidirectional path through the hepatocytes (Hershko et al, 1979). Ferrioxamine formed extracellularly does not readily enter hepatocytes and is eventually excreted in the urine (Keberle, 1964; Hershko etal, 1979). DFO was introduced as a chelating agent in the 1960's (Keberle, 1964). Because it is poorly absorbed from the gastrointestinal tract, DFO must be administered parenterally. In 1964 Smith et al, (1964) showed that a slow continuous intravenous infusion of DFO markedly increased urinary iron excretion, this mode of drug administration was subsequently considered impractical. Interest was sparked in the 1970's when several researchers reported that intramuscular DFO administered 6 days per week significantly increased urinary iron excretion, decreased the rate at which serum ferritin levels increased and inhibited both hepatic iron accumulation and tissue damage (Modell et al, 1974; Risdon et al, 1975). In 1977, Propper et al, (1977) described the use of a small portable battery operated infusion pump to administer DFO subcutaneously, which is still in use today for the treatment of iron overload. This mode of administration proved to be 90% as effective as an equivalent intravenous dose in terms of urinary iron excretion. Without doubt, regular chelation therapy with DFO has been shown to prolong life expectancy, reduce liver iron and establish a negative iron balance (Giardina et a l , 1985, 1990; Olivieriet al, 1994). Patients compliant with DFO are less likely to develop cardiac complications than patients who are not compliant or who start chelation therapy at an older age. Intensive chelation therapy with intravenous DFO has been shown to improve cardiac function in some patients with pre-existing disease concomitant with a rapid decline in serum ferritin levels, an increase in urinary iron excretion and improved complience (Cohen, 1990). DFO has been shown to improve, endocrine function. In one study sexual maturation did not seem to correlate with the intensity of chelation therapy. However, in another study when chelation therapy was started before puberty there was an increased chance of achieving normal sexual maturation (Bronspiegel-Weintrob et al, 1990). Indeed early use of DFO has been 43 G H c - G - N / \ / \ (C H2)2 ( 0 ^ ) 5 H3 H2H(H2C)5 yX (9 ^ ) 5 ^ ( ^ ^ 2)2 N— C n - a N-G / / / V HO G HG G HO Figure 1.10: Structure of Desferrioxamine CK CH 3 Figure 1.11: Structure of CP20 44 shown to decrease the risk of endocrinopathies such as diabetes melhtis (Brittenham et al, 1994). Despite the obvious benefits of chelation therapy with DFO in thalassaemia patients, there are still significant concerns regarding its safety, particularly at high doses. Regimens in which the dose exceeded 50mg/kg administered intravenously over 1 2 hours in well chelated patients with minimal iron burden has been associated with a variety of toxicities eg. neurological, pulmonary, renal, bone and growth abnormalities (De Virgilis et al, 1988; Freedman et al, 1988; Gabutti et al, 1990; Koren et al, 1991). Because the effectiveness of DFO is greatly enhanced by continuous infusion, DFO is generally administered subcutaneously over an 8-12 hour period 5 to 7 days a week by means of a small portable infusion pump. However this causes many problems which limit its usage. It is highly expensive and inconvienient to use because it is orally inactive and it only causes sufficient iron excretion to keep pace with transfusion regimes. For this reason many patients, particularly adolescents find it difficult to comply with the treatment regime. While DFO presents these problems, it is nonetheless the standard with which any new iron chelator must be compared, since it can achieve negative iron balance in iron overloaded patients and manifests relatively few side effects. Unfortunately DFO treatment is not universally available and there are at least 30,000 children born each year with thalassaemia, mostly in Asia, for whom sub-cutaneous treatment with portable infusion pumps is neither practical nor affordable (Hershko, 1988). Therefore an inexpensive, orally active, iron chelator is highly sought after. The obvious method of choice when designing new iron chelators is to model compounds on the structures of naturally occurring siderophores, a growth promoting agent secreted by microorganisms to scavenge iron from the environment. Siderophores are low molecular weight compounds with a very high affinity for iron. Coordination of the metal occurs in most cases via the oxygen atoms of either hydroxamates or catechol moieties. However hydroxamates like DFO as stated earlier are susceptible to the acid environment of the stomach. An approach presently under investigation is chemical attachment of DFO to a hydroxyethyl starch polymer (Mousa et al, 1992) which then creates a high molecular weight chelator with affinity for iron identical to that of desferrioxamine but with a vascular half-life 10 to 30 times longer. Therefore single infusions of this compound may improve compliance with long-term chelating therapy. The expense and inconvienience of DFO has led to a 25 year search for an orally active chelator and a number of candidates have been tested in the past (Porter et a/.,1989, Hershko and Weatherall, 1988). Two major obstacles stand in the way of the development of effective oral substitutes for DFO, these are the availability of appropriate animal models for drug testing and the possibility of ensuring that candidate 45 chelators are given the degree of toxicological screening that is appropriate for a drug that will be continuously used throughout the life of the patient. Initially animal testing should use a simple reproducible cheap system in which several compounds can be compared at the same time. Once a compound has been found to produce significant iron excretion, the results should be confirmed in a second animal species. This is particularly important as basal iron excretion and the proportion of iron chelated into faeces compared with urine are higher in laboratory animals than in humans (Finch et al, 1978). Many years ago, Pippard et al, 1981 developed a rapid screening model in non-iron-overloaded rats using a single pulse of ^^Fe ferritin to label hepatic parenchymal cells followed by a challenge with a test chelator 2 hours later, a time when iron released by lysosomal degradation of ferritin is maximally available. Although extracellular chelation of RE iron by DFO in urine was thought to be influenced by the degree of iron overload in the rat, the model was able to identify the urinary excretion of cold iron with DFO (Pippard et al, 1981). Clinical trials of any drug should follow adequate toxicity testing for the appropriate length of time in two species of animals as laid down by the various drug regulatory authorities. Important considerations for the design of new iron chelating drugs with regard to toxicity screening have been summarised in a recent review (Hider et al, 1996), as it is almost impossible to construct a chelator for Fe(II), which does not interact with other divalent transition metal cations such as Cu(II), Co(II), or Zn(II). Chelators for Fe(III) are likely to be much more selective. Among the candidates most worthy of consideration are derivatives of the bidentate hydroxpyridin- 4-one series (Porter, 1989) of which CP20 (LI, Deferriprone) is an example. 1.5.2.2 Hydroxypyridinones The hydroxypyridinones were designed by Hider and colleagues (Hider and Silver, 1982). These compounds are orally active due to their neutral charge in both the iron bound and the iron free form and their high chemical stability which protects against attack by digestive enzymes in the gastrointestinal tract. Most attention has focused on the l,2-dimethyl-3-hydroxypyridin-4-one derivative (CP20, Deferriprone, L I) (Figure 1.11). CP20 is a bidentate ligand, therefore, three molecules are required to form a neutral complex with one atom of Fe(III), compared with the hexadentate ligands such as DFO, which binds in a 1:1 ratio with Fe(III). Furthermore CP20 forms a very stable iron complex (log p3=35.7) which has been shown not to catalyse the formation of hydroxyl radicals (Singh and Hider, 1988) unlike other hydroxypiridinones which have either a 2 : 1 or 1 :1 coordination with iron and may catalyse hydroxyl radical formation because of free coordination sites (Kontoghiorghes, 1995) 46 CP20 is rapidly absorbed and reaches a peak plasma level at a mean of 45mins (Matsui et al, 1991). It is metabolised to a glucuronide and is excreted in the urine as follows: 1. unchanged, 2. bound to iron, 3. as its glucuronide derivative and, 4. bound to trace metals such as zinc and aluminium. The efficacy of the drug in heavily iron- loaded patients, assessed by the amount of the drug excreted in urine bound to iron in 24 hours compared with the amount of a single oral dose, has been estimated to be approximately 4% (Al-Refaie gr a/., 1995). This study also showed that the amount of iron excreted in the urine in heavily iron-loaded thalassaemia major patients after a single oral dose was significantly related to the time area under the curve for CP20 in plasma. The sites from which CP20 chelates iron are as yet not established. Studies in normal and iron-loaded rats with radioactively labelled CP20 have shown it concentrates mainly in the liver (Hileti et al, 1995). Because free CP20 readily enters cells, it is likely that both parenchymal and reticulo-endothelial cells are sources of chelated iron. Unlike DFO, CP20 can also chelate iron from transferrin and Al-Refaie et ai, (1995) estimated that up to 20% of iron excreted in the urine after a single oral dose may be delivered from iron bound to transferrin. It has also been shown that CP20 chelates iron from intact red cells which may be an important source of iron that is excreted in response to CP20 therapy in thalassaemia intermedia (Shalevet al, 1995). Over 300 thalassaemic patients worldwide are on long term CP20 therapy. In transfused patients with thalassaemia major, 75mg CP20 per Kg body weight induces urinary iron excretion approximately equivalent to that achieved with 30-40mg of DFO per Kg sufficient to induce negative iron balance in many patients with thalassaemia major, as documented by decreasing serum ferritin and by liver biopsies (Kontoghiorghes, 1987; Olivieri et al, 1990; Brittenham, 1992). Furthermore CP20 has been shown to remove hepatocellular iron via the bile and causes urinary excretion of the reticuloendothelial system iron in experimental animals (Zevin et a l, 1992). CP20 is especially suitable for patients who cannot tolerate DFO because of anaphylactic reactions, ophthalmic, auditory, pulmonary and neurological toxicity. As with long term DFO, CP20 therapy carries the risk of side effects. Indeed because faecal iron excretion induced by CP20 is much less than that by DFO, the short term efficacy of CP20 is inferior to that of DFO (Zevin et al, 1992; Collins et al, 1994). The lipophilicity of the hydroxypyridinones enables them to penetrate numerous cells where unfortunately they can exert toxic effects. Toxicity studies of CP20 in non- iron-loaded animals have reported many unwanted reactions including anaemia, leucopenia, bone marrow and thymic atrophy, growth retardation and embryotoxicity. In humans, the first reported toxic effect of CP20 was agranulocytosis in a woman with Blackfan Diamond anaemia (Hoffbrand et al, 1989). Further episodes of agranulocytosis or of severe neutropenia have occurred in several other patients and 47 despite severalin vitro studies, the mechanism for the agranulocytosis remains obscure. The second most important adverse effect is joint toxicity (arthropathy), first described by Bartlett et al, (1990). The syndrome consists of musculoskeletal stiffness, joint pains and, in severe cases, joint effusions. Other complications include nausea and zinc deficiency. Concerns regarding the adverse effects of CP20 on immunologic function were raised in a case report describing fatal "systemic lupus erythematosus" in a patient receiving CP20 in India (Mehta et al, 1991) in studies reporting inhibition of human lymphocyte proliferation by CP20 in vitro and in studies describing thymic atrophy in rats (Berkoukas et a l, 1993). Indeed results from a recent randomised trial of CP20 and DFO has raised concerns that long-term therapy with CP20 may not provide adequate control of body iron in a substantial proportion of patients with thalassaemia major (Olivieri et al., 1995). Despite the toxicity studies on CP20, it has been licensed for clinical use in India and currently it seems likely that 60-70% of patients with thalassaemia major and other transfusion dependent refractory anaemias will be able to take CP20 long term without unacceptable side effects and with sufficient efficacy to prevent an increase in iron stores. Although many alternative potentially orally active iron chelators have been tested in tissue culture, in animal models and in short term clinical trials, none have as yet proved to be sufficiently safe and effective for their consideration for long term clinical trials. 1.5.2.3 Clinical consequences of chelator therapy Iron chelators are a unique family of drugs whose safe action is limited to the iron-loaded state, promoting the excretion of excess potentially toxic iron. In normal or iron deficient subjects all iron chelators are toxic. Foremost among such toxic effects is inhibition of ribonucleotide reductase, an enzyme responsible for the reduction of ribonucleotides to deoxyribonucleotides, a rate limiting step in DNA synthesis. The enzyme contains a redox-active iron at its active centre. Its removal by agents such as DFO and CP20 has been proposed to underlie the anti-proliferative effects of such agents. DFO has been shown to have anti-proliferative effects in many cell types including normal (Hoffbrand et ai, 1976) and malignant cells (Lederman et a i, 1984). As previously outlined the main toxicities of CP20 at this time are, bone marrow aplasia and thymic atrophy in animals (Porter et ai, 1991) and neutropenia, agranulocytosis and liver enzyme abnormalities in humans. Although the mechanism of neutropenia is unknown, in some cases it is clearly an immune reaction to CP20, as illustrated by the simultaneous development of agranulocytosis, systemic vasculitis, alterations in humoral and cell-mediated immune function and the presence of circulating immune complexes (Castriota-Scandenberg & Sacco, 1997). Despite several in vitro studies the mechanism for agranulocytosis remains obscure (Al-Refaie et a i , 48 1994, Cunningham et al, 1994). Like DFO, CP20 also inhibits DNA synthesis by chelating iron from ribonucleotide reductase (Hoyes et al, 1992) and this may well be a cause of agranulocytosis. It has been shown that CP20 causes anaemia, with increased corpuscular volume and leucopenia in normal and iron-loaded mice and rats (Porteret a l, 1991). It is unclear whether these effects occur by a similar mechanism to the idiosyncratic neutropenia or agranulocytosis in humans. Porter et al, (1994) and Hileti et al, (1995) suggested that the clinical consequences of both DFO and CP20 may be explained by the phenomenon of apoptosis and showed that the chelators, DFO and CP20 promoted apoptosis in murine thymocytes and human leukaemic cells. 1.6 Apoptosis Apoptosis is a mode of cell death in which single cells are deleted in the midst of living tissue. The term derives from a Greek word used for the dropping off of leaves from trees. It is characterised by structural changes that appear with great fidelity in cells of widely different lineage and presumably represent a pleiotropic effector response (Kerr, 1972; Arends, 1991). Apoptosis is the descriptive name given to the process of programmed cell death and is essential for the appropriate development and function of multicellular organisms. Unnecessary, damaged and potentially harmful cells must be deleted from surrounding healthy cells to ensure structural and functional tissue homeostasis. In adult tissues, cell death figures prominently within tissues such as the endometrium, prostate, adrenal and mammary glands, it is also fundamental to the development, regulation and function of the immune system. This includes the elimination of self-reactive thymocytes, negative selection of B and T lymphocytes and cell killing effected by cytotoxic T lymphocytes. Apoptosis is an evolutionarily conserved, innate process by which cells systematically inactivate, disemble and degrade their own structural and functional components to complete their own demise and death. It can be activated intracellularly through a genetically defined developmental program or extracellularly by cytokines, hormones, xenobiotic compounds, radiation, oxidative stress and hypoxia. The ability of a cell to undergo apoptosis in response to a death signal is related to its proliferative status, cell cycle position, and the controlled expression of genes that promote, inhibit, and affect the death program. Stringent regulation of these death modulating parameters must be maintained to ensure that apoptosis occurs in the proper physiological context (Webb et al, 1997). 49 1.6.1 Morphological characteristics of apoptosis Apoptosis characteristically affects scattered single cells, not groups of adjoining cells. The process can be divided into three distinct stages: 1. Commitment, in which the cell having received a potentially fatal stimulus becomes irreversibly committed to death. 2. Execution, during which the major structural and biochemical changes occur and 3. Degradation, when the cellular remnants are cleared away by phagocytes. The earliest morphological changes are compaction and segregation of the nuclear chromatin, with the formation of delineated granular masses that become marginated against the nuclear envelope, and condensation of the cytoplasm. Molecular characterisation of the chromatin reveals an ordered degradation of the DNA by a cation-dependent nuclease (Wyllie et al, 1980), first to large fragments of 30-50 kilobases (Oberhammer er a/., 1993) and finally into nucleosomal fragments of 180- 200 base pairs (Wyllie, 1980). Although many studies of cell death have suggested that the morphological changes associated with apoptosis occur as a result of DNA laddering, others have shown that morphological cell death can occur in the complete absence of intemucleosomal fragmentation (Cohen et al, 1992; Oberhammer et al, 1992). Therefore the precise role of DNA fragmentation in apoptosis remains controversial. The progression of the DNA condensation is accompanied by a convolution of the nuclear and cell outlines. The stmctural integrity of the plasma membrane is further compromised by a loss of the phospholipid asymmetry, micovilli and cell-cell junctions. The cell rounds up, dissociates from its neighbours, shrinks dramatically and produces buds of membrane-bounded 'apoptotic bodies'. Apoptotic bodies are quickly ingested by neighbouring cells and are degraded by their lysosomes. The apoptotic process is extremely rapid, typically occurring after the initiating stmulus between a few minutes and a few hours and the apoptotic debris is cleared with similar rapidity (Kerr, 1972). 1.6.2 Biochemical characteristics of apoptosis Distinct biochemical changes are known to occur during apoptosis. One of the most prominent alterations is the fragmentation of the DNA into 180-200bp oligomers, resulting in what is known as a 'DNA ladder'. However observations by several investigators suggest that large DNA fragments (30-50 and 200-300kb) and even single strand cleavage events occur during apoptosis (Cohenet al, 1994; Solis-Recendez et a l, 1995). Large fragments have been reported to occur both in the presence and absence of oligonucleosomal fragments (Oberhammer et al, 1993). The large fragments may not be precursors of the oligonucleosomal fragments as both types of fragments can be produced, apparently independently, under some conditions (Bortner et al, 1995). The characteristic fragments of apoptosis were originally thought to be 50 produced by double strand cleavages as a result of frequent nicks on both strands of DNA in the intemucleosomal regions (Peitsch et ai, 1993). Recent evidence however has shown that these breaks are more specific, producing double stranded DNA fragments that are either blunt ended or have 3' single base overhangs with hgatable 5 ' ends. The endonuclease responsible for DNA cleavage during apoptosis has eluded identification, but several candidates have been proposed Each nuclease has unique properties differing in tissue distribution, ionic dependency and pH optimum (Cotter, 1995). 1. DNase (I): which requires Ca^+ and Mg^+ and is inhibited by zinc. Antibodies against DNase 1 and specific inhibition of DNase 1 reduce the intemucleosomal DNA cleavage activity characteristic of apoptosis in thymocyte extracts. 2. DNase (II): This nuclease can cause the characteristic intemucleosomal DNA fragmentation pattem of apoptosis in isolated CHO cells. It has an optimum pH of 5.5 and can function independently of divalent cations. The predominant subcellular location of DNase (II) is in acidic lysosomes, therefore given the low pH at which this enzyme functions and its subcellular locale it does not seem like a good candidate. 3. NUC 18: This apoptotic enzyme is dependent on Ca^+ and Mg^+, is also inhibited by zinc and has a nuclear localisation. 4. CAD: Recently Enari et al, (1998) have identified a factor that can cleave chromatin at intemucleosomal sites. The nuclease is expressed in many cell types, usually in the cytoplasm in an inactive, latent form. The group found that it can be activated by Caspase-3 digestion, earning it the name, Caspase-activated DNase (CAD). Evidence is accumulating that cytoplasmic structures such as mitochondria participate in the critical effector stage of apoptosis (Kroemeret al, 1995; Mignotte and Vayssiere, 1998), with investigators suggesting that mitochondria contribute to apoptosis signalling via the production of reactive oxygen species (ROS). The addition of ROS or the depletion of endogenous antioxidants can promote apoptosis (Ratanet al, 1994; Sato et al, 1995; Guenal et al, 1997). Furthermore, increases in intracellular ROS are sometimes associated with programmed cell death (PCD) (Martin and Cotter, 1991). However, the nature of ROS in apoptosis is a conflicting question and it has been difficult to ascertain that the observed accumulation of ROS corresponds to a causal effect and is not simply one of the side effects involved in the killing process. Several changes in mitochondrial biogenesis and function are however associated with the commitment to apoptosis. A fall in mitochondrial membrane potential (A'Pni) was shown to occur before the nuclear changes of fragmentation of the oligonucleosomal fragments (Vayssiere et al, 1994; Zamzami et al, 1995). The A'Fm is the result of the asymmetric distribution of charges between the inner and outer 51 mitochondrial membranes. Thus, an ion-impermeable membrane and the function of the respiratory enzyme complexes that pump protons out of the matrix space are required for the generation and maintenance of an intact A'Fni- The A'Fm is necessary for the establishment of an electrochemical proton gradient and, consequently, for mitochondrial ATP synthesis. Reduction of A'Pm correlates with a decrease in mitochondrial gene translation, loss of mitochondrial gene transcription and an uncoupling of the respiratory chain from oxidative phosphorylation (Wolvetanget al, 1994). A reduction in A'Prn and associated apoptosis occurs in many cell types, such as lymphocytes exposed to glucocorticoids, activation of peripheral T cells, T hybridomas and pre-B cells (Zamzami et al, 1995), in neurons deprived of nerve growth factor (NGF) and during apoptosis of U937 and HeLa cells induced by TNF- a . Another biochemical event known to occur during apoptosis is cytoplasmic acidification. Cytoplasmic acidification is believed to be essential for apoptosis by activating pH dependent enzymes (Meisenholderet al, 1996). Concomitant with this are changes occurring to the plasma membrane. Phosphatidylserine (PS), which is normally restricted to the inner layer of the bilayer, becomes externalised in cells committed to apoptosis (Fadok et ai, 1992). This exposure allows the early recognition and phagocytosis of apoptotic cells. Although PS is currently the only known qualitative difference between the plasma membranes of apoptotic and viable cells (Martin et al, 1995; Koopman et al, 1994), several studies have shown that macrophages use other ways of recognising dying cells. Activated macrophages use a lectin-like receptor to recognise apoptotic cells, whereas resting macrophages use an integrin vitronectin receptor (Schlegelet al, 1996). 1.6.3 Apoptosis vs Necrosis Until the elucidation of apoptosis, necrosis was the classically defined form of cell death. A cell exposed to a stress, will either survive, undergo apoptosis or necrosis, the choice depends on the level of stress. Necrosis as a form of cell death is however completely distinct from apoptosis (Wyllieet al, 1980) (Figure 1.12). Necrosis results from either direct damage to the plasma membrane by agents such as complement or by interference with energy dependent ion pumps. The result of this is an intake of water and the cell swells, which is in contrast to the cell shrinkage seen in apoptotic cells. Other characteristics of necrosis are: 1. Increase in mitochondria density as the outer membrane swells. 2. Flocculation of nuclear chromatin 3. Decrease in protein synthesis 4. Cytoskeletal disruption and the release of lysosomal enzymes 52 o LOW ^ INSULT A ► 'HIGH ^ m INSULT ► O o I Heat shock i^proteins /^EDIU NECROSIS UNSULT A p o p to sis program APOPTOSIS Figure 1.12: Apoptosis vs Necrosis A ceil exposed to stress will either survive, undergo apoptosis or necrosis. The choice depends on the level of stress. 53 5. Random DNA fragmentation 6 . A fall in pH Throughout the process of necrosis the nuclear membrane remains intact. However, the cellular contents are released into the extracellular space and can invoke an inflammatory response (Wyllie et al, 1980). Finally it is known that apoptosis is genetically regulated, which is the pivotal difference between the two modes of death. 1.6.4 Apoptosis and disease In multicellular organisms, homeostasis is maintained through a balance between cell proliferation and cell death. In most tissues, cell survival appears to depend on the constant supply of survival signals by neighbouring cells and the extracellular matrix. Evidence suggests that the failure of cells to undergo apoptosis may contribute to the pathogenesis of a variety of human diseases, including cancer, autoimmune disease and viral infections as outlined below in Table 1.4 (Thompson, 1995). However a wide number of diseases characterised by cell loss, such as neurodegenerative disorders, AIDS (acquired immunodeficiency syndrome) and osteoporosis, may result from accelerated rates of apoptosis. Diseases associated with the Diseases associated with inhibition of apoptosis increased apoptosis 1. Cancer: 1. AIDS Follicular lymphomas 2. Neurodegenerative Carcinomas with p53 mutations D isorders: Breast/Prostate/Ovarian cancer Alzheimers disease 2. Autoimmune disorders: Parkinsons disease Systemic lupus erythematosus Retinitis pigmentosa Immune-mediated glomerulonephritis Cerebellar degeneration 3. Viral infections: 3. Ischaemic injury Herpesvirus Heart Attack Poxvirus Stroke Adenovirus Reperfusion injury Table 1.4: Disorders associated with apoptosis Adapted from Thompson, 1995 A number of effective methods that induce target cells to undergo apoptosis already exist. Both chemotherapeutic agents and radiation induce tumour cell death by 54 causing damage that induces the cell to commit suicide (Fischer, 1994). Conversely, treatments that increase a cell's resistance to undergo apoptosis may be of benefit even in the absence of specific alterations in the genes involved in cell death regulation. Current evidence suggests that the susceptibility of cells to undergo apoptosis is regulated continuously. Thus, treatments that can increase the apoptotic thresholds of specific cells may be beneficial in the disorders associated with cell loss. Examples of this include the use of growth factors to promote haematopoietic cell recovery after cancer chemotherapy and treatments with antioxidants to prevent the death of CD4+ T cells in response to HIV infection. Several observations suggest that the central mediators of apoptosis may still be pharmacologically manipulated in a cell-specific fashion, for this to occur an understanding of the genetics of apoptosis is required. 1.7 MOLECULAR BIOLOGY OF APOPTOSIS 1.7.1 Role of Caenorhabditis elegans The nematode caenorhabditis elegans {C. Vegans) is a very powerful model system to study the mechanisms of apoptosis. The embryonic development of C. Elegans is reproducible and has been precisely mapped. 1090 somatic cells are eventually formed in the adult hermaphrodite and of these 131 undergo programmed cell death (Ellis et a l, 1991). This developmentally programmed cell death shows several areas of similarity with apoptosis in mammalian cells. Mutant nematodes have been identified with defects in different parts of the cell death process and this has allowed a genetic pathway of cell death to be produced (Miura et al., 1996; Yuan, 1996). The genes involved can be divided into several groups: genes involved in triggering cell death; those involved in the cell death process itself; those required for engulfment and finally those involved in the disposal of the cell corpse. These genes are known as Ced genes (cell death genes). Three genes have been shown to affect the execution of cell death, Ced-3, Ced-4 and Ced-9. The activity of two of these genes, Ced-3 and Ced-4 is required in cells for death to occur. The third gene, Ced-9 is required to protect cells that should survive from undergoing apoptosis, this gene encodes a protein that is homologous to the Bcl-2 family of cell death regulators (Webb et al, 1997) (Figure 1.13). An important advance came with the discovery that the death effector, Ced-3 had significant homology with the mammalian protease interleukin-Ip converting enzyme (ICE) (Yuan et al, 1993). ICE is involved in the proteolytic cleavage of the inactive 31 kD alL -ip precursor at Asp^^^/Ala^^^ to release the 17.5 kDa mature form of the active cytokine. Originally isolated from monocytes, its unique function was 55 Engulfing cell Decision Execution to die of death Engulfment Degradation Healthy cell Healthy cell Dead cell a i committed to die O) ced-9 ced-1 nuc-1 ced-2 ced-5 ced-6 ced-3 ced-7 ced-4 ced-8 ced-10 All dying cells Figure 1.13: Ced genes involved in apoptosis emphasised by the fact that its genetic sequence did not resemble that of any other known protease (Thornberry et al, 1992; Cerretti et al, 1992). Recently, a series of biochemical studies have provided a clearer understanding of the molecular interactions of the components of the cell death apparatus. Using yeast two-hybrid analysis of ectopic expression in mammalian cells , several groups have demonstrated that Ced-4 can bind to Ced-9 (and human Bcl-xL), Ced-3, or both simultaneously (Chinnaiyan et al, 1997; Spector et al, 1997; Wu et al, 1997 and Cryns et al, 1998). These interactions are functionally important for cell death signalling, for instance, Ced-9 or human Bcl-xL mutants that are defective in their ability to inhibit cell death do not bind to Ced-4. In addition Ced-4 binding to Ced-3 is mediated by a conserved amino acid terminal protein interaction module, the so called Caspase recruitment domain (CARD) (Irmler et al, 1997; Zou et a l, 1997). Ced-4 binding to Ced-3 via this domain in the presence of ATP promotes the autoproteolytic activation of Ced-3 and its ability to induce apoptosis (Seshagiri and Miller, 1997). Importantly, Ced-9 blocks the ability of Ced-4 to activate the Ced-3 killer protease, thereby preventing cell death. Ced-9 may also act downstream of Ced-3 as a pseudosubstrate inhibitor of this protease (Xue and Horvitz, 1997; Cryns et al, 1998). 1.7.2 Role of proteases in apoptosis In mammals there are currently 11 different homologues of Ced-3 identified. These 11 different cysteine proteases with aspartate specificity are now called caspases for Cytosolic Aspartate-Specific cysteine Proteases.(Table 1.5) (Alnemri et al, 1996; Wang gf al, 1998). The Ced-3/ICE-like proteases/Caspases share the following structural and functional characteristics: 1. They all induce apoptosis when overexpressed; 2. they are all synthesized as proenzymes (Figure 1.14) and require proteolytic processing to form the active enzyme,3. they share the same conserved active site consisting of the pentapeptide sequence QACRG and4. they all cleave their substrates after an aspartate residue. In addition to their divergent substrate specificities, caspases differ in the length and sequence of their amino terminal domains, caspases -1,-2,-4,-5,-8,-9 and-10 all have long prodomains, whereas caspases -3, 6 -, -7 and-11 have short prodomains. Two distinct protein-protein interaction modules have been identified in the long prodomains. The first has been called the death effector domain (DED), two copies of which are present in caspases - 8 and -10 (Boldin et al, 1996; Vincenz and Dixit, 1997). This domain targets caspases - 8 and -10 to ligand activated cell death receptors such as Fas/APO- 1/CD95 via specific protein interactions with a DED module in the adapter protein FADD/MORTl (Boldin et al, 1995; Chinnaiyan et al, 1995). In contrast the prodomains of caspases -1, -2, -4 and -9 all contain a caspase recmitment domain, CARD, which as detailed earlier can also be found in Ced-4 (and its 57 mammalian homologue Apaf-1) and the death adapter protein RAIDD/CRADD (Duan and Dixit, 1997; Zou et al, 1997). This domain was subsequently shown to mediate the interaction between Apaf-1 and pro-caspase 9, a necessary event leading to caspase 9 and subsequently caspase-3 activation. Therefore the presence of a family of these structurally related caspases in cells raises questions in relation to their roles in cell death eg. 1. Are all caspases required for cell death or are some members more important than others? 2. Do all modes of death utihse the same caspases? 3. Are different caspases activated in series or in parallel?4. Are all caspases tissue specific? It is evident from studies such as those with caspase-1 and caspase-3 knockout mice that not all caspases are required for cell death. Caspase-3 deficient mice are smaller than their littermates and die at 1-3 weeks of age and thymocytes from caspase- 3 deficient mice show a similar sensitivity to apoptosis as normal thymocytes induced to die by a number of different stimuli including dexamethasone. Brain development in these deficient mice is however markedly affected with a variety of hypoplasias being observed from embryonic day 12 (Kuida et al., 1996; Cohen, 1997). Asp 4 APOPTOSIS Figure 1.14: The caspase cascade The proenzyme is activated by cleavage at specific Asp residues (1), which allows the subunits to re-order as an active heterodimer (2). The heterodimer can activate other pro-enzymes (3), which results in a caspase cascade leading to cell death (4). 58 Figure 1.14, illustrates how the ICE-like proteases (caspases) function in apoptosis. The caspases are synthesized as pro-enzymes, which are cleaved at an Asp residue to produce a small and a large subunit that associate as heterodimers. This active form of the enzyme has the ability to cleave and activate other pro-caspases, producing a protease cascade that results in the death of a cell. As suggested earlier overexpression of ICE or Ced-3 in mammahan cells induces apoptosis which suggests that ICE, or a related protease, is a component of the cell death pathway. Unlike Ced-3 in the nematode, ICE is not necessary for all the apoptotic events in mammalian cells. Indeed it was shown that mice with the null genotype for ICE develop normally, suggesting that ICE is not required for cell death in the developing mouse embryo and that other proteins with similar activities must exist (Kuida eta/., 1995; Li etal, 1995). 1.7.3 Substrates cleaved by caspases during apoptosis Several proteins important in many of the cellular processes affecting the cell such as DNA repair, cytoskeletal integrity and cell signalling have been shown to be proteolytically cleaved during apoptosis, many are substrates of the caspase family of proteases. These proteins include, poly ADP-ribose polymerase (FAR?), DNA-PK, the cytoskeletal proteins, lamins A, B and C, fodrin, gas2 and actin; the cell signalling molecules, PKCô and the U1 small nuclear ribonucleoprotein particle polypeptide U l- 70. Some caspases have overlapping specificities for substrates (caspase-3 and caspase-7 can both cleave PARP) whereas other caspases may have a unique substrate specificity (caspase- 6 is the only known caspase to cleave lamins). PARP, a key enzyme in DNA repair and in the supervision of genome structure and integrity in stressed cells (Nicholson et al, 1995) is one of the best characterised substrate of caspases, being cleaved in the execution phase of apoptosis in many systems, including, thymocytes and HL-60 cells. Intact PARP is a 116 kDa protein which is cleaved to 24 kDa and 89 kDa fragments in virtually every form of programmed cell death examined. The cleavage of PARP is an early event of apoptosis and the loss of its function coincides with endonuclease activation. PARP is cleaved at the sequence DEVD*G by a protease resembling ICE, but not ICE itself (Lazebnik et al, 1994). The protease which cleaves PARP was independently identified by three research groups (Fernandes-Alnemri etal 1994; Tewari et al, 1995; Nicholson et a l, 1995) and has been designated caspase-3. Although the cleavage of PARP is a valuable indicator of apoptosis, its relevance is unclear, since PARP-null mice develop normally (Leister a/., 1997). The proteolysis of lamins, the major structural proteins of the nuclear envelope, is observed in many cells undergoing apoptosis and may be responsible for the nuclear 59 changes that occur, since inhibitors of lamin cleavage prevent some of these changes (Lazebnik et al, 1995; Neamati et al, 1995). Caspase- 6 cleaves lamin A at a conserved VEID*N sequence. The activity of the U1 small nuclear ribonucleoprotein particle, which is essential for the splicing of precursor mRNA, is dependent on both the RNA and protein components, including Ul-70. In several systems including CD95, Fas- induced apoptosis, Ul-70 is cleaved early in the apoptotic process to a 40kDa fragment (Casciola-Rosen et al, 1996, Tewari et al., 1995). Cleavage of Ul-70 separates the RNA binding domain from the distal arginine-rich region of the molecule which may have a dominant negative effect on splicing; such inhibition of splicing would block cellular repair pathways dependent on new mRNA synthesis (Casciola-Rosenet al, 1996). As outlined in section 1.6, during the execution phase of apoptosis, a number of important plasma membrane changes occur resulting in the recognition and phagocytosis of the apoptotic cell by a phagocyte or a neighbouring cell. Cleavage of important cytoskeletal proteins including fodrin (Martin et al, 1995; Vanags et a l, 1996; ). a-fodrin, an abundant membrane-associated cytoskeletal protein is rapidly and specifically cleaved during Fas-induced apoptosis and this appears to be related to membrane blebbing (Cryns et al, 1996). 1.7.4 Inhibitors of Caspases Further evidence supporting a critical role for ICE-like proteases in apoptosis is the ability of specific protease inhibitors, including the cowpox viral serpin CrmA (Miura et al, 1993, Gadliardini et al, 1994; Tewari et al, 1995) and baculovirus p35 (Clem, 1991), to inhibit apoptosis. Use of synthetic oligopeptide inhibitors designed to mimic the recognition/cleavage sites of known Ced-3/ICE substrates in cell free apoptotic assaysin vitro has further substantiated the role of the caspases in apoptosis (Table 1.5). One of the most common synthetic irreversible inhibitors is Z-VAD-fmk [benzyloxycarbonyl-Val-Ala-Asp (OMe) fluoromethyl ketone]. Z-VAD-fmk has been shown to be a potent cell permeable inhibitor of apoptosis in a number of different systems, such as thymocytes, Jurkat T-cells and hepatocytes (Jacobson et al, 1996; Armstrong etal, 1996). Z-VAD-fmk inhibits apoptosis at a very early stage, as shown by its inhibition of many of the ultrastructural features of apoptosis including PARP cleavage (Slee etal, 1996; Jacobson et al, 1996). Recently several investigators have shown that it inhibits the processing of caspases -2, 3, 6 , 7 and 8 , suggesting that it inhibits a caspase at or near the apex of the caspase cascade (Macfarlane et al, 1997). 60 P rotease Alternative names Inhibitors Caspase-1 ICE YVAD, DEVD, p35 lAP, WEHD, crmA Caspase-2 ICH-1, NEDD-2 lAP, p35 Caspase-3 CPP32, Yama, Apopain DEVD, (crmA???) Caspase-4 ICErel-II, XX, ICH-2 YVAD, DEVD Caspase-5 ICErel-III, TY Caspase- 6 Mch2 VEID Caspase-7 Mch3, ICE-LAP3, CMH-1 DEVD, p35 Caspase- 8 FLICE, MACH, Mch5 crmA Caspase-9 ICE-LAP6 , Mch6 Caspase-10 Mch4, FLICE2 DEVD Caspase-11 Table 1.5: The human caspase family and their inhibitors 61 1.7.5 A caspase hierarchy Using a combination of inhibition studies and the fact that the proteases seem to be processed at aspartate residues, together with their ability to cleave at aspartate residues suggests that some of the proteases activate themselves or act upon each other (Figure 1.14) and therefore, it is now possible to begin to map out the pathway of caspase activation. The main focus has been on caspase-3 which has been shown to cleave PARP, it appears to be the central mediator in most studies on apoptosis and directly processes pro-caspases 2,6 , 7 and 9 (Srinivasula et ai, 1996), conversely many caspases have been shown to activate caspase-3, including caspase-1. Two potentially interacting cascade initiating pathways converge on the activation of the downstream effector caspases that act to kill the cell by cleaving death substrates. The first of these pathways is initiated in response to apoptotic stimuli such as DNA damaging agents that trigger the mitochondrial release of cytochrome c into the cytoplasm; once in the cytoplasm, cytochrome c interacts with other factors to form a caspase-3 activating complex (Yang et al, 1997; Cryns et al, 1998). In the second pathway, caspase proenzymes are recruited to ligand bound death receptors via homophilic interactions with adapter proteins thereby leading to the proteolytic activation of these upstream caspases and initiation of the cascade. A system in which this process of programmed cell death is well understood, indeed one of the most investigated systems, is that of the CD95 (Fas/Apo-1) receptor and its relatives. Fas is a type I membrane protein belonging to the tumour necrosis factor (TNF)/nerve growth factor (NGF) receptor family (Itohet al, 1991). The Fas ligand (Fas L) is a member of the TNF family and is a type II protein. The binding of FasL to the Fas receptor has been shown to induce apoptosis in Fas-bearing cells (Itoh et al, 1991; Nagata and Goldstein 1995). The Fas receptor and TNF-receptor l(TNFRl) both contain a conserved cytoplasmic sequence that has been termed the death domain (DD) (Nagata., 1997). The DD is required for initiating the apoptosis process inside the cell because a single point mutation in this domain abrogates the apoptotic signal. Subsequent to the identification of the DD in Fas and TNFRl, several other proteins containing homologous death domains have been isolated. These include. Reaper (a drosophilia protein) (Whiteet al, 1994), F ADD (Fas-associated Jeath domain)/ MORT (Mediator of receptor-induced toxicity) (Boldin et a l, 1995; Chinnaiyan et al, 1995), TRADD (TNF-receptor associated Jeath d/omain) (Hsu et al, 1995), RAIDD (FlP-associated /ch- l/Ced-3-homologous protein with <7eath Jomain)/CRADD (caspase and RIP adapter with Jeath domain) (Duan and Dixit, 1997), MADD (Schievella etal, 1997) and RIP (receptor interacting protein) (Stranger et al, 1995). Overexpression of any of these proteins in mammalian cells induces apoptosis. 62 Caspases 8 and 10 contain FADD-like death domains, suggesting that they may be components of a Fas-mediated apoptotic pathway and as such are believed to act at the top of the apoptotic pathway to activate a protease cascade (Fernandeset ai, 1996) (Figure 1.15). Activation of these initial caspase proteases could subsequently activate downstream members of the caspase family such as caspase-3,6 and 7 which could then cleave proteins observed to be cleaved during apoptosis. Other types of receptor-mediated activation may involve other caspases such as caspase 10. It has been proposed that caspase 10 may be involved in many CrmA-insensitive non-receptor mediated cell death as it is poorly inhibited by CrmA (Srinivasula et al, 1996). 63 TNFa FasL Fas TNFRl Grov/th Factor Withdra'.va! Irradiation, Drugs TRADD FADD FADD ? Aggregation? h - ^ 3 - Ecl-2 _H Family ?Caspas 9 Caspase 8 Caspase 10 H3 SH HS Caspase 3 Caspase 6 CrmA p35 Caspase 7 PAR? U170K DNA PK Lamins. etc Ceil Death Figure 1.15: The caspase Cascade Apoptosis can be induced by a number of different stimuli including growth factor withdraw!, irradiation treatment with various drugs, or stimulation of the Fas or TNFRl receptors. Following Fas receptor activation the adapter protein FADD binds to the death domain (DD) found on the cytoplasmic tail of Fas by means of its own C-terminal DD. The N-terminal FADD domain (hatched oval) can bind to one or more proteins containing similar domains such as those found on caspase-8 or caspase-10. Activation of these initial caspase proteases could subsequently activate downstream caspase members such as caspase-3, -6 and -7 which could then cleave the proteins observed to be cleaved during apoptosis. 64 CHAPTER 2 General Materials and Methods 65 2.1 General materials and methods * 0.5, 1.5ml polypropylene tubes Eppendorf, Germany * 25ml universal tubes Sterilin, Feltham,UK * 50ml polypropelene tubes Falcon, Becton Dickinson, Oxford * 5, 10, 25ml pipettes Sterilin, Feltham,UK * Tissue culture flasks Falcon, Becton Dickinson, Oxford * Pipette tips Gilson Anachem, Luton * Phosphate Buffered Saline (PBS) Gibco BRL, Paisley, UK * Dexamethasone (DFX) Sigma, Poole, UK * Hydroxyurea (HU) Sigma, Poole, UK 2.2 Iron chelators Desferrioxamine (DFO, Desferal) was obtained as a lyophilised mesylate from Novartis (Basle, Switzerland). It was made up as a tenfold concentrated stock in PBS and added to cells, 1:10. The hydroxyp3|i'idinones, 1,2-dimethyl-3-hydroxypyridin-4-one, (CP20, LI or Deferriprone) supplied as a white powder was synthesized as previously described (Dobbin et al., 1993) at the Department of Pharmacy, Kings College London and the purity confirmed by elemental analysis IH NMR, and reversed phase HPLC. CP20 was made up as a tenfold concentrated stock in PBS and added to cells 1:10 to achieve the appropriate final concentration. Hexadentate chelators such as DFO coordinate iron in a 1:1 ratio, whereas the bidentate 3-HP-4-ones coordinate in a 3:1 ratio. Hence three moles of HPO's bind the equivalent amount of iron to one mole of DFO. The iron complexed forms of the chelators were prepared by the following procedure: 1. Make up stock of ferric chloride (mol wt. 270.3), lOmM in 0.1 M HCL 2. Make up stock of chelator, eg. CP20 (mol wt. 175.5), lOOmM in 10ml PBS 3. To the chelator stock, add 3mL of iron solution, bubble through with nitrogen to prevent oxidation and quickly bring the pH up to 7.4. This results in a final iron: chelator molar ratio of 1:3 for the preparation of the bidentate 3-HP-4-one iron complexes. 66 2.3 Cell Culture Materials * RPMI 1640 containing L-glutamine Gibco BRL, Paisley, UK * Foetal Calf Serum(FCS) Gibco BRL, Paisley, UK (heat inactivated for 30mins) * Penicillin/Streptomycin (Pen/Strep) Gibco BRL, Paisley, UK * Bovine Serum Albumin (30%) Sigma, Poole, UK * Ethidium bromide Sigma, Poole, UK * Fluorescein-di-acetate Sigma, Poole, UK * Trypan blue Sigma, Poole, UK 2.3.1 Cell Lines The following established cell lines were used: HL-60 derived from a patient with acute promyelocytic leukaemia. Collins et al, 1977 Method All cell culture studies were performed in sterile plastic tissue culture flasks. All cell lines were maintained in suspension culture in RPMI 1640 supplemented with 10% FCS and 2% Pen/Strep in a humidified atmosphere containing 5% CO2 at 37°C and subcloned twice a week to a density of 0 .6 - 1 x 1 0 ^ cells/ml to ensure that exponential growth was maintained. Cell number was determined manually by microscopic counting using an improved neubûu^e^counting chamber. Cell viability was measured using an ethidium bromide/fluorescein-di-acetate viability stain. 2.3.2 Cell counting and viability Cell number was determined manually by microscopic counting using an improved neubeauer counting chamber. Cell viability was measured using an ethidium bromide/fluorescein-di-acetate viability stain. This was prepared by adding 4ml of ethidium bromide (4mg/ml) to 10ml of fluorescein-di-acetate (0.1% solution) and this solution was made up to 1 litre with PBS. The solution was stored in the dark at 4°C until required. Viability was assessed by mixing 100>Aof cell suspension with 100/t/l of ethidium bromide/fluorescein-di-acetate stain. The percentage of viable cells was counted using a 67 fluorescent microscope. A minimum of 500 cells were counted for each sample. Ethidium bromide selectively enters dead cells which then fluoresces an orange-red colour, whilst, fluorescein-di-acetate is taken up by viable cells which appears a green colour under UV light (Mills etal, 1983). 2.3.3 Freezing and thawing of cell lines For freezing, cells were suspended at 5-10x10^ cells/ml in 10% dimethyl sulphoxide (Sigma, Poole, UK) and 20% FCS in RPMI medium. Aliquots of 1ml were frozen overnight in cryotubes at -70°C and then transferred to liquid nitrogen for storage. Frozen cells were rapidly thawed in a 37®C waterbath and diluted by dropwise addition of RPMI 1640 medium supplemented with 10% FCS and 2% Pen/Strep to 20ml. Cells were washed once, resuspended and counted before use. 2.3.4 Cytospin preparations Single cell suspensions were prepared at 1.0x10^ cells/ml and 100//U of cell suspension was added to each prepared funnel with labelled slides and filter attached. The centrifuge (Cytospin 2, Shannon) was spun at 700 rpm for 10 minutes. Slides were removed, air dried and stained with May Grunwald Giemsa staining for 30 minutes 2.3.5 Thymocyte Isolation Thymocytes were obtained from male Balb-C mice and aged 10-12 weeks, unless otherwise stated. Following cervical dislocation, the thymus was carefully dissected and contaminating connective tissue removed. The thymocytes were released from the capsule by gentle teasing apart the thymus in culture medium (RPMI 1640 medium, supplemented with 10% FCS and 2% Pen/Strep), debris was removed from the suspension via filtration through a plastic gauze. The cells were suspended at a concentration of 1.0x10^ cells/ml in culture medium and incubated in 5ml polypropelene tubes at 37°C in 5% CO2 . Cells were counted using a haemocytometer and viability determined by trypan blue exclusion. Cells of a >95% purity were used for subsequent experimentation. 2.3.6 Culture of bone marrow cells from murine bone marrow Male C57bL/J mice aged 10-12 weeks were sacrificed by cervical dislocation. The femur was isolated and the haemopoietic cells were flushed from the bone cavity with 5mis of RPMI medium supplemented with 10% FCS and 2% Pen/Strep, using a 25G hypodermic needle. The isolated cells were washed x2 with 5mis of the same medium and 68 then resuspended at a concentration of 1x10^ cells/ml. After 24h incubation at 31^C with the drugs of interest, cells were pelleted and fixed in 70% ethanol for 60mins. After this time, cells were washed x3 with 3% BSA/PBS, a rat anti-mouse 1^ neutrophil antibody (Serotec, Oxford, UK) was added at a dilution of 1:4 and left on ice for 30mins. Cells were once again washed x3 with 3% BSA/PBS to get rid of any free antibody. The 2^ antibody, a F(ab')2 Rb anti-rat IgG(AFF.PUR): FITC (Serotec, Oxford, UK) was added at a dilution of 1:50 and again left on ice for 30mins. After several washes, cells were resuspended with 0.25ml of 3% BSA/PBS, 0.25ml RNase, 1 mg/ml and 0.5ml propidium iodide, 50pg/ml. Cells were left in the dark at 4^C until analysis on the FACS for % of apoptosing cells as oulined in section 2.4.1 2.3.7 Isolation of Peripheral blood progenitor cells Patients with haematological malignancies were mobilised with either cyclophosphamide (1.5g/m^) and G-CSF (263 pg daily) or ESHAP chemotherapy (etoposide 40mg/m^ for 4d, cA-platin 25mg/m^ over 4d, cytarabine Ighrfi over 4d and methylprednisolone 500mg daily) and G-CFS. CD34+ cells were then isolated from the leukopheresis product by one of two methods. In some cases the entire hai'vest product was positively selected using the Ceprate SC stem cell concentration system and the 12.8 biotinylated anti-CD34 antibody as previously described (Watts et al, 1996). Alternatively, a small aliquote of the leukapheresis product was removed and the CD34-I- cells were positively selected using the Variomacs TM system (Miltenyi Biotech) as per manufactures instructions. Briefly, mononuclear cells were isolated by Ficoll-Flypaque deASity centrifugation. The cells from the interface were washed and incubated with lOOpl of multisort magnetic beads conjugated to the anti-CD34 monoclonal antibody Q-BEND/10 for 30 minutes on ice then selected on a variomacs column. Product purity was determined by APAAP (alkaline phosphatse anti-alkaline phosphatase) for the CD34 antigen on cytospins as previously described (Wattset al, 1996). 69 2.4 Measurement of apoptosis 2.4.1 Flow Cytometric quantification of apoptosis Materials * Phosphate Buffered Saline (PBS) Gibco BRL, Paisley, UK *Ethanol Analar, UK * RNase (Type lA) Sigma, Poole, UK * Propidium Iodide (PI) Sigma, Poole, UK Method Quantitive comparison of apoptosis was obtained by measuring the proportion of hypodiploid DNA present in whole cells, (Nicolettiet al, 1992). Cells were centrifuged at 1500 rpm for 10 minutes and the supernatant removed. The centrifuged cell pellet was fixed in 1ml of cold 70% ethanol at -20®C for 30min-24h. The cells were then centrifuged, washed x2 in 1ml PBS and resuspended in 0.25ml PBS. To a 0.25ml cell suspension, 0.25ml RNase (1 mg/ml) was added followed by the addition of 0.5ml PI (50pg/ml in PBS). The mixed cells are incubated in the dark at 4°C until measured.Cells were analysed using a cytometer cell sorter (EPICS Elite, Coulter Electronics Ltd, Luton, UK). The cells were illuminated at 488nm and the instrument was set to detect forward, 90^C light scattering (side scatter) and fluorescence at 575nm. A low level discriminator eliminated the contribution from very small particles and electrical noise. The results shown represents the data collected from at least 1 0 , 0 0 0 cells and nuclei contributing to the sub-diploid distribution were scored as being apoptotic. 2.4.2 Agarose Gel Electrophorhesis Materials Apoptosis Buffer 1ml lOmMEDTA 20/Alof500mMEDTA 50mM Tris (pH 8.0) 5 0 ^ of IM Tris 0.5% Sodium lauryl sarkinosate 50yUl Of 10% Nalsar 0.5mg/ml Proteinase K (PK) 50./Aof lOmg/ml PK 8 3 0 ^ of DD water 70 lOx TBE Running buffer(pH 8.3) IL Tris 108.9g Oithoboric acid 55.7g EDTA, Disodium salt 7.4g Loading Buffer 20mM Tris(pH 8.0) 10ml Glycerol 10ml B romophenol Blue 0.01% Method 1x10^ cells were spun at 2000 rpm for 5mins, pelleted cells were resuspended in 20|il of apoptosis buffer (containing 500mM EDTA, IM Tris, 10% Sodium Lauryl Sarkinosate and lOmg/ml proteinase K) and incubated at 4^C overnight. After incubation at 50°C for Ih, RNase (0.5mg/ml) was added. Incubation continued for Ih and the samples warmed to 70°C. Loading buffer, lOpl was then added and the DNA loaded into dry wells of a 2% agarose gel containing 0.1 mg/ml ethidium bromide. The samples were run in Ix TBE electrophoretic buffer at lOOV until the marker dye had migrated approximately 4cm. DNA laddering was visualised using a UV transilluminator (Smithet al, 1989) 2.5 Measurement of intracellular zinc concentrations Materials *Hanks Buffered Salt Solution(HBSS) Gibco BRL, Paisley, UK *Zinc Sulphate (ZnS04) Sigma, Poole, UK *Digitonin Sigma, Poole, UK *Zinquin AMRAD Corp. Australia Method 1x10^ thymocytes/ml in culture medium (RPMI containing L-glutamine and supplemented with 10% FCS and 2% Pen/strep) previously treated for 24h with various concentrations of Zinc-Sulphate (0-200jiM) were incubated with zinquin (final concentration, 25jiM) in HBSS for 30 minutes at 37^>C. To obtain a spectrophotometric measurement, the fluorescence of unloaded cells (due to autofluorescence and light scattering) was subtracted from the readings to derive zinquin dependent fluorescence. In 71 some experiments, washed cells were lysed in the cuvettes by addition of digitonin, SOjiM and the released Zinquin was saturated with 25pM Zn SO 4 , to devive Fmax. Fmin was devived by further addition of IM HCL to quench Zn-dependent fluorescence. Fluorescence was measured at room temperature in a Perkin-Elmer LS 50 luminescence spectrophotometer. Single excitation and eirussion spectra peaks were observed at wavelen^s of 370nm and 490nm respectively Fluorescence readings by spectrofluorimetry were converted into pinol of Zn/10^ cells using a standard curve (Figure 2.1) derived by titration of increasing amounts of ZnSOq into a solution of 3pM zinquin, until the fluorescence was equivalent to that obtained with zinquin labelled cells. The medium for the titration was a solution with an ionic constitution resembling that of the lymphocyte cytosol (125mM 20mM N»C ImM Mg“"^, Hepes buffered to pH 7.05) and supplemented with 0.1 mg/ml BSA (Tsien et a i, 1982). 300 1 W c 3 200 - 0 ) Ü c 0 ) 0) 0) 100 - 0 3 0 H—...... —I I IIIIIII—n Tmn|—i iiiiiii|—i i i iiiif—r 11 itm] ,001 .01 .1 1 1 0 1 00 1 000 [ZnS04] nM Figure 2.1: Zinquin standard curve 72 2.6 Determination of Haemopoietic subsets Materials Control jri IgG-ECD (Coulter, Luton, UK) 200 PC5/PE/FITC (Coulter, Luton, UK) 400 Tube 1 CD7-PE/CD2-FITC (Coûter, Luton, UK) 200 CD19-ECD (Coulter, Luton, UK) 200 CD 10-Tricolour (Coulter, Luton, UK) 200 Tube 2 CD33-FITC (Coulter, Luton, UK) 200 CD13-PE (Coulter, Luton, UK)* 200 *CD13-PE stock is diluted 50|il into 150|li1 PBS. Then 200)il of this diluted antibody is used to make the cocktail. Method CD34+ cells isolated as indicated in section 2.3.7, were cultured in RPMI-1640 medium with lOng/ml SCF, IL3 and IL6 . To investigate the expression of surface antigens, CD13, CD33 (Myeloid), CD7, CD2, CDIO and CD19 (Lymphoid) and CD3 4 , cc sample containing 1x10^ cells/ml was spun at 3000rpm for 3 minutes. After this time, the supernatant was decanted and the sample resuspended in SOOpl PBS (pH 7.4). To lOOpl of sample, 5pi of either the control, tube 1 or tube 2 antibody cocktail was added and incubated for 20 minutes at room temperature. After this time, 500/4 of PBS was added. Samples were analysed by flow cytometry as indicated in section 2.4.1. 2.7 Statistical Analysis In this thesis the Student T test has been employed to statistically analyst the data. The T test is used for small samples and assumes that the populations are normal 73 CHAPTER 3 Apoptotic Induction in Various Cell Types by Iron Chelators 74 3.1 Introduction. In Section 1.5 of Chapter 1, it was stated that hydroxypyridinone iron chelators and in particular l,2-dimethyl-3-hydroxypyridin-4-one (CP20, deferiprone or LI) have been observed to induce bone marrow hypoplasia and thymic atrophy in laboratory animals (Porter et al, 1991; Hoyes et al, 1993) and agranulocytosis in a variable proportion of humans receiving CP20 for the treatment of transfusional iron overload (Al-Refaie et al, 1995). These effects have not been generally observed with the iron chelator in most common clinical use, desferrioxamine (DFO), either in laboratory animals or in iron overloaded humans. Porter et ai, (1994); Fukuchi et al., (1994) and Hileti et al, (1995) suggested that these effects of could be explained by the process of apoptotic induction by iron chelators in both thymocytes and HL60 cells. Apoptotic induction by iron chelators has also recently been described in a variety of cell lines including the human leukaemic CCRF-CRM cell line (Haq et al, 1995) and the mouse B-cell lymphoma cell line 38C13 (Kovaret al, 1997). In this chapter, the susceptibility of thymocytes and bone marrow progenitors to apoptotic induction by both the hydroxypyridinone chelator, CP20 and Desferrioxamine (DFO) has been studied. This was in order to determine to what extent apoptotic induction by chelators is a unique feature of hydroxypyridinones as opposed to a general feature of iron chelators. Thymocytes were chosen because of the previously reported thymic atrophy with hydroxypyridinones (Porter et al, 1993, 1994). Purified human haemopoietic CD34+ cells were chosen because the bone marrow is an important potential target organ for CP20 and because these cells provide a convenient model to study the effects of iron chelators on haemopoietic cells of the myeloid lineage as they mature under culture conditions in vitro. Previous work by Hoyes et a l, (1993) showed that the iron chelators CP20 and DFO have potent anti-proliferative effectsin vitro on haemopoietic colony formation but the mechanism involved and stage(s) of cellular differentiation at which CP20 exerts its effect remain unclear. Apoptosis could in principle be a consequence of the anti-proliferative actions of iron chelators as it plays a key role in the homeostatic control of many haemopoietic cell types. Recently, it has become apparent that apoptosis is just as fundamental to maintaining cellular balance in terms of tissue development and control of cell numbers as cell proliferation. Cells inherently die unless continuously signalled not to do so and the immune system and the control of haemopoiesis offer many examples of this theory in practice (Allen et al, 1993). For example, the myeloid leukaemic cell line HL60 undergoes apoptosis after in vitro differentiation following exposure to retinoic acid (Martin et a l, 1990) and peripheral blood neutrophils and eosinophils undergo 75 apoptosis when maintained for prolonged periods of time in culture (Cotteret al., 1994). At the time of isolation CD34+ cells are predominantly quiescent (Pardee, 1989; Burke et ai, 1992; Lemoli et al, 1997; Williams et al, 1997) but during in vitro culture in the presence of appropriate growth factors they subsequently proliferate and differentiate into mature myeloid types, thereby allowing a systematic examination between the effects of cell proliferation, differentiation and apoptosis. 3.2 Apoptotic induction by iron chelators in thymocytes 3.2.1 Rationale Apoptosis is well known to play a central role in the development and clonal deletion of thymocytes. Extensive cell death occurs during thymocyte maturation, when up to 90% of immature T cells are deleted by apoptosis during the processes of positive and negative selection that shape the lymphocyte compartment (King and Ashwell, 1993). Several years ago it was recognised that thymocytes which recognise self antigen with a high affinity may mature into autoreactive lymphocytes and therefore must be deleted (clonal deletion), this is negative selection (White et a l, 1989). More recently, clonal deletion has been shown to arise through the specific induction of apoptosis (Murphy et al, 1990; MacDonald et al, 1990) Thymocytes undergo a stringent selectivity process and only those cells which recognise self antigen with a relatively low affinity are allowed to mature into cells expressing both CD4+ and CDS"*" (positive selection). Numerous signals can trigger apoptosis in thymocytes and several of these have been well characterised such as the response to glucocorticoids and low dose gamma irradiation (Sellins and Cohen, 1987; Ishii et al, 1997; Stefanelli et al, 1997). Following the observation that hydroxypyridinones induced thymic atrophy in rats given repeat doses in vivo (Porter et al, 1993), it was further found that iron chelators can also promote the induction of apoptosis in thymocytesin vitro (Porter et al, 1993, 1994; Maclean gr a/. 1995, 1996, 1997). In this chapter results are presented on studies investigating the kinetics of apoptotic induction in thymocytes by the iron chelators, CP20 and DFO in order to determine whether differences in the rates of thymocyte apoptosis in vitro could explain the apparent lack of thymic atrophy within vivo use of DFO. 3.2.2 General Experimental procedures Murine thymocytes were dissected from the thymuses of male Balb-C mice aged 10-12 weeks and cultured as indicated in Section 2.3,5. The cells were resuspended at a concentration of 1x10^ cells/ml in RPMI medium containing L-glutamine, supplemented with 10% FCS and 2% Pen/Strep. The iron chelators CP20 and DFO were used at a concentration of 300|liM IBE/ml unless otherwise stated (see sectio n 76 2.2) and dissolved in phosphate buffered saline (PBS). In each experiment, an equivalent volume of PBS was used as a negative control and, for experiments using murine thymocytes, the glucocorticoid Dexamethasone (DEX) (10"^M) was used as a positive control of apoptosis. Apoptosis in thymocytes was demonstrated using four techniques; light microscopy, agarose gel electrophoresis, quantitation of the hypodiploid DNA peak by flow cytometry, and immuno-analysis of the weak staining CD4+CD8+ apoptotic population by flow cytometry. R esu lts 3.2.3 Demonstration of apoptotic features by light microscopy. The morphology of thymocytes was assessed for the presence of apoptotic cells using light microscopic examination of May Grunwald-Giemsa cytospin preparations (section 2.3.4). Morphological observations of thymocytes, fresh or treated for 24h with CP20, BOOpM EBB using light microscopy revealed the characteristic morphological changes associated with apoptosis after 24h incubation with CP20. Freshly dissected thymocytes appeared as small round cells with a high nuclear to cytoplasmic ratio, the nuclear chromatin was dense and homogeneous and they were non-granular (Figure 3.1a). However after 24h incubation with either PBS or the iron chelator, CP20 the morphology of the cells changed significantly (Figure 3.1b), with marked condensation of their chromatin, nucleolar disintegration and a shrinkage in cell volume with concomitant increase in cell density compared to control freshly isolated thymocytes. These changes were consistent with those previously described for thymocytes undergoing apoptosis (Wyllieet al., 1980; Arends etaL, 1991). Time-lapse videomicroscopy permits observation of the onset and execution of individual apoptotic events within a cell culture. As previously discussed, the onset of apoptosis is characterised by the sudden initiation of surface blebbing, together with cytoplasmic fragmentation and exfoliation. Evan et al., (1992) showed that in fibroblastic cells, the onset of blebbing is rapid and is quickly followed by nuclear condensation, collapse and total cell fragmentation, a process that was shown to take between 30-60 min (Evan et at., 1992; McCarthy et al., 1997). Indeed Cohen et al., (1984) showed that apoptosis begins to be detectable by light microscopy within one or two hours of in vitro incubation with dexamethasone. 3.2.4 DNA fragmentation by agarose gel electrophoresis Agarose gel electrophoresis has been widely used to visualise the DNA fragmentation characteristic of apoptosis in a variety of cells including thymocytes (Smith et al., 1989; Sun et al., 1992; Cohen et al., 1994; Bortner et al., 1995). The 77 e 0# # H # ## B * Figure 3.1: Normal (A) and apoptotic (B) thymocytes stained with May-Grunwald Giesma 78 details of the methodology used are described in section 2.4.2. Briefly thymocytes treated for 24h with either PBS, CP20 and DFO (300|iM IBE) or DEX at 37®C/5%C02 were centrifuged and resuspended overnight at 4°C in 20|Li1 apoptotic buffer [500mM EDTA, IM Tris, 10% sodium lauryl sarkinosate (NaLSar) and lOmg/ml proteinase K]. Samples were then incubated at 50®C for Ih, after which RNase, 0.5mg/ml, was added and incubation continued for a further Ih at 50^C. Samples were warmed to 70°C and lOjil loading buffer added (10ml of 20mM Tris, pH 8.0, 10ml glycerol and 0.01% BPB) for 30 minutes. Samples were loaded into dry wells of a 2% agarose gel containing 0.1 mg/ml ethidium bromide. Samples were electrophoresed in a Ix TBE electrophoretic buffer at lOOV for 45min and the DNA visualised on a UV transilluminator. The DNA fragmentation characteristic of the intemucleosomal cleavage seen with apoptosis and resulting in DNA laddering on agarose gel electrophoresis is shown in Figure 3.2 for thymocytes incubated for 24h at 37^C with DFO or CP20. DNA from freshly isolated thymocytes showed little or no DNA laddering (Lane 2), but following 24h incubation at 37®C in culture medium significant DNA laddering was seen (Lane 3). This laddering was also present in CP20 treated cells (Lane 4) and DFO treated cells (Lane 6). DNA laddering with Dexamethasone at 24h (Lane 7) was used as a positive control. The pattern of DNA laddering with dexamethasone was similar to that previously reported in thymocytes (Bortner et al., 1995) and did not differ qualitatively from the pattern observed with the iron chelators DFO or CP20 after 24h. Furthermore, previous work from our group showed that the iron chelators and dexamethasone could demonstrate DNA laddering as early as 6h incubation (Lynagh et a/., 1994) Recently, however several research groups have detected large DNA fragments of 50-300kb during apoptosis. Brownet at., (1993) and Cohen et at., (1994) both showed that rat thymocytes induced to undergo apoptosis by treatment with dexamethasone demonstrated not only intemucleosomal DNA cleavage, as seen in Figure 3.2, but also the presence of large DNA fragments. Cohen et al., (1994) showed that after Ih incubation of thymocytes with DEX, large DNA fragments before the appearance of DNA laddering, which occurred after 4h incubation could be detected. These results suggested that the formation of large fragments of DNA preceded endonuclease cleavage of DNA into nucleosomal fragments. The problem with microscopy and agarose gel electrophoresis as methods of detecting apoptosis is that the results are qualitative and not quantitative. 3.2.5 Quantitative analysis of DNA fragmentation using flow cytometry In order to obtain quantitative comparisons of the rate and concentration dependence of apoptosis induced by iron chelators, flow cytometric analysis of the 79 1. Marker 2. Fresh 3. 24h PBS 4. 24h CP20 5. 24h HU 6. 24h DFO 7. 24h DEX Figure 3.2. Apoptosis detection using Agarose Gel Electrophoresis Agarose Gel Electrophoresis of DNA extracted from mouse thymocytes after 24h incubation using freshly extracted thymocytes or thymocytes incubated for 24h with either PBS, the iron chelators CP20, DFO (300pM IBE), HU(lmM) or DEX ( 10-%). 80 hypodiploid peak of DNA obtained from thymocyte nuclei was used. This is an established method for quantification of thymocyte apoptosis (Nicoletti et a l, 1992; Sun et al., 1992). To quantify the amount of apoptosis a treated cell suspension of 1x10^ cells/ml was centrifuged and resuspended in 70% ethanol at -20°C until required (minimum 30min incubation). Before analysis the cells were then washed 3x in PBS and resuspended in 0.25ml PBS, 0.25ml RNase (Img/ml) and 0.5ml PI (50|ig/ml). Cells were mixed and incubated overnight at 4°C. Cells were analysed using a cytometer cell sorter (EPICS Elite) (section 2,4.1). 10,000 cells were collected for analysis and those contributing to the sub-diploid distribution were scored as being apoptotic (see section 2,4,1). Figure 3,3 shows the flow cytometric profiles obtained from fresh thymocytes or thymocytes treated for 24h with either the negative and positive controls, PBS or DEX (10"^M) respectively, or the iron chelator, CP20 (300)liM IBE). The profiles represents the DNA content of cells with apoptosis were scored as sub-diploid. When murine thymocytes were incubated with either PBS, CP20, DFO, 300|iM IBE or DEX (lO'^M), the percentage of cells demonstrating apoptosis, as represented by the sub diploid distribution, was significantly increased in CP20-treated cells (59.8+4.5%) at 24h compared to control PBS treated cells (34.9±4.1%), p<0.001. There was a small but significant increase in apoptosis in DFO-treated cells at this time compared to control (44.2+5.5%), p<0.05, and dexamethasone treated cells also showed an increase in apoptosis (70.9+3.2%), p<0.001 (Figure 3,4). Viability loss was quantified using fluorescein di-acetate and ethidium bromide staining as indicated in section 2,3,2. In CP20- and DFO-treated cells, the percentage of non-viable cells after 24h incubation was 30.1+5.2% and 31.5+3.2% respectively compared to PBS, 21.4±2.4%. Hence apoptotic induction could be detected in chelator- treated cells before loss of viability could be detected. 3.2.6 Identification of apoptosis by immunoanalysis of CD4+8+ th ym ocytes Apoptosis is the mechanism by which both autoreactive and unselected immature CD4^8^ thymocytes are eliminated in the thymus. Using flow cytometric analysis to detect changes in the level of surface expression of CD4 and CDS molecules of thymocyte cultures in suspension it was possible to identify apoptosis of CD4^8^. For detection of apoptosis of immature CD4"*"8+ thymocytes, the rat monoclonal antibodies, FITC-anti-CD8+ and PE-anti-CD4+ (Sigma, Poole, UK) were used at optimal concentrations (1:4 of stock solution) determined by antibody titration and staining, and washing of cells was done in cold PBS (4^C) containing 5% FCS. For staining, 1x10^ murine thymocytes from 10-12 week old male Balb-C mice in RPMI culture medium containing L-glutamine, supplemented with 10% FCS and 2% 81 Fresh % Apoptosis 24h DEX Figure 3.3: Effect of iron chelators on apoptosis: Flow cytometric profiles Thymocytes were freshly isolated or incubated for 24h with controls, PBS or DEX ( lO'^M) or the iron chelator CP20, 300pM IBE. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown is the typical flow cytometric profile obtained with addition of the drugs. 82 8 0 - CN 60- OD W 40- O w Q. O a 2 0 - < 0 - PBS CP20 DFO DEX Figure 3.4: Effect of iron chelators on thymocyte apoptosis after 24h incubation Thymocytes were incubated with either controls, PBS or DEX (10" M) or 3(X)pM IBE CP20 or DFO for 24h. The cells were washed, fixed in 70% ethanol and then stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 8 independent experiments done in duplicate 83 Pen/Strep, were pre-incubated for 24h with PBS, CP20, DFO or Dexamethasone at 37°C in a humidified atmosphere of 5% CO2 in air. This was then further incubated for 30 minutes at 4®C in the mixture of optimally diluted PE-anti-CD4"*" and FITC-anti- CD8+ antibodies (1:4 dilution) in a final volume of 0.2ml. Negative control cells were either unstained or contained either CD8^ or CD4^ antibodies only. After this time cells were washed 2x and resuspended with SOOpl PBS prior to analysis on EPICS CS cytometer as indicated in section 2.4.1. Investigation of the expression of CD4 and CD8 molecules by flow cytometric analysis revealed that fresh murine thymocytes were 19.3% CD4+CD8, 16.4% CD8^CD4, 4.5% CD4 CD8- and 59.8% CD4+CD8+. After 24h CD4+CD8+ thymocytes cultured for 24h with either PBS, CP20 or DFO (300|iM IBE) or DEX expressed lower levels of CD4^ and CD8^ (circled populations 1 and 2) than freshly isolated thymocytes (Figure 3.5). The murine CD4+8"'' thymocytes expressing lower than normal levels of CD4 and CD8 molecules represents a sub-population of immature thymocytes undergoing apoptosis (Swatet al., 1991). After 24h in culture PBS, CP20, DFO or DEX the percentage of cells in population 1 was 49.8%, 24.8%, 30.4% and 9.9% respectively and for population 2 was 8.4%, 29.9%, 27.2% and 49.7% respectively for PBS, CP20, DFO (300)liM IBE) and DEX respectively. The formation of CD4+CD8+ thymocytes expressing reduced levels of CD4 and CD8 molecules after in vitro culture has been observed by other investigators (Inaba et al., 1988; Kyewsky et al., 1989), but previously none of these studies identified such cells as apoptotic. On the contrary, they were assumed to represent a viable population of cells and the degree of cellular DNA was not tested. However in 1991, Swat et al., (1991) demonstrated complete degradation of DNA as measured by agarose gel electrophoresis isolated from this sub-population, thereby concluding that the weakly stained CD4^CD8^ population represented apoptosis. 3.2.7 Time dependence of chelator induced apoptosis To investigate the effect of time on thymocyte apoptosis, 1x10^ cells/ml in culture medium were incubated with CP20 or DFO at 300|liM IBE for 0, 2, 4, 8 or 24h, then treated for analysis by flow cytometry. To investigate the effect of a short pulse of iron chelators on thymocytes, cells were incubated for 4h with either PBS, CP20, DFO or DEX. After this time the samples were washed 3x in culture medium, resuspended in 1ml of culture medium and the incubation at 37°C allowed to continue for a further 24h prior to analysis by flow cytometry. Table 3.1 shows the effect of the duration of exposure of thymocytes to chelators. Significant differences between the chelators and control, PBS were seen as early as 4h, with maximal differences at 24h, p<0.05. 84 3.5a Fresh Thymocytes 3.5b 24h Ctrl F:1T2 LB FHT2 LC6 3.5c 24h CP20 3.5d 24h DEX ic 1?0 m . PMT2 LC« 6iz.t Figure 3.5: Expression of CD4+CD8+ thymocytes in suspension culture Thymocytes (1x10^ cells/ml) were freshly isolated or incubated for 24h with controls, PBS or DEX (lO'^M) or the iron chelator CP20, 300pM IBE .The surface expression of CD4 and CD8 molecules on cell populations was analysed after in vitro culture by direct staining with the mixture of PE-labelled anti-CD4 and FITC-labelled anti-CD8 monoclonal antibodies. Groups 1 and 2 represent non-apoptotic and apoptotic cell populations. 85 Time (h) Control CP20 DFO DEX 0 2.0 ± 0.3 00 a i 4 4.8 ± 1.5 14.1 ± 3.2** 9.8 ± 2.7* NA 8 20.3 ± 5.7 35.5 ±6.6* 33.8 ± 3.5* NA 24 34.9 ±4.1 59.8 ± 4.5** 44.2 ± 5.5* 70.9 ± 3.2*** Table 3.1: Effect of incubation time on chelator-induced thymocyte apoptosis Thymocytes were incubated for 0, 4, 8 or 24h with controls, PBS or DEX or the chelators, CP20 and DFO, 300p,M IBE. The cells were washed, fixed in 70% cold ethanol and then stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. *, ** and *** denotes significance of p<0.01, p<0.05 and p<0.001 respectively as measured by the students T test. Similar results were also obtained when investigating the effect on thymocyte apoptosis with a short pulse, 4h, of iron chelators. Results again showed an increased amount of apoptosis after 24h with PBS (37.3±2.0%) compared to freshly isolated thymocytes (2.0±0.3%). Similarly the iron chelators increased the amount of apoptosis (57.9±1.9% and 46.6+1.9% for CP20 and DFO respectively, p<0.05), comparable to that occurring after a continuous exposure of 24h. The difference between CP20 and DFO with regard to their propensity to cause an increase in the amount of apoptosis in thymocytes may be explained by the chemical nature of CP20 compared to DFO. CP20, as outlined in Section 1.5 has a lipophilic nature and would be predicted to have greater access into cells and organelles than hydrophilic compounds like DFO. Furthermore CP20 has a much smaller molecular weight than DFO thereby allowing CP20 to exert its effects on the cell and cellular organelles at a much faster rate than DFO. Viability staining (section 2.3.2) of thymocytes incubated with PBS, chelators or DEX for either 0, 2, 4, 8 or 24h showed that the percentage of freshly isolated non- viable thymocytes was 12.6+3.9%. Viability was only significantly decreased between chelator-treated cells and dexamethasone-treated cells compared to control after 8h continual incubation (24.3+4.1%, 28.9+6.1% and 14.7+1.5% non-viable cells for CP20, DEX and PBS-treated cells respectively, P<0.05). Therefore once again apoptotic induction could be detected in chelator-treated cells before loss of viability could be detected. 3.2.8 Concentration dependence of chelator induced apoptosis The effect of chelator concentration was examined by incubating cells, IxlO^/ml in culture medium for 24h at 37°C/5%C02 with 0, 11, 33, 100, 300 or 900|iM IBE CP20 or DFO, prior to subsequent analysis for apoptosis by flow cytometry as outlined in section 2.4.1. Figure 3.6 shows the dose response effect of either CP20 or DFO on the percentage of thymocytes apoptosing after 24h incubation with the iron chelators as measured by flow cytometry. Increased apoptosis compared to control at 24h was observed at concentrations as low as llpM IBE, with CP20 having maximum effect at 300)liM (64.3±1.9%), in contrast to DFO which achieved maximal apoptosis at 900|iM IBE (58.8+2.4%), p<0.05. 3.2.9 Discussion Thymocytes have been widely recognised as an important model for studying apoptosis, since apoptosis now represents the main mechanism of intrathymic cell selection (see introduction). A variety of stimuli such as corticosteroids, calcium ionophores and anti-CD3 monoclonal antibodies are well known to induce apoptosis in 87 801 CP20 DFO S' 40: Q_ i 0 200 400 600 800 1000 Chelator Concentration (//M IBE ) Figure 3.6: Effect of chelator concentration on thymocyte apoptosis at 24h Thymocytes were incubated for 24h with chelator concentrations of, 0, 11, 33, 100, 300 and 900pM IBE. After 24h the cells were washed, fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 88 thymocytes incubated in vitro. (Wyllie et al., 1980; Sellins et al., 1987; Smith et al.; 1989; WalkQvetal., 1991). In this study thymocyte apoptosis was investigated in thymuses from 10-12 week old Balb-C mice, since this was found to be the optimum age at which apoptosis occurred (see section 2.3.5). Indeed, a recent paper by Provinciali et al., (1998) showed a significant decrease of both the total number and the proportion of CD4^CD8^ double positive thymocytes present in old (21-22 months) and very old mice (24-26 months) in comparison with young animals. As stated above, well known agents for inducing apoptosis in thymocytes are glucocorticoids. Recently, Ishii et a i, (1997) suggested that the majority of thymocytes exposed to glucocorticoids in vivo die without exhibiting evidence of DNA fragmentation. Using the terminal deoxynucleotidyl transferase (TdT)-mediated dUTP- biotin nick end-labelling (TUNEL) technique, they showed that flow cytometric analysis did not detect a single TUNEL+ thymocyte, even after 4h treatment with glucocorticoids. This suggested that virtually no free dead thymocytes were present after DNA fragmentation and that typical apoptosis induced by glucocorticoids which is characterised by DNA fragmentation is not the dominant type of cell death in the murine thymus. This hypothesis would seem to be contradictory to that presented in this study where apoptotic DNA fragmentation in the thymus following the addition of both glucocorticoids (DEX) and iron chelators was clearly evident (Figure 3.2). However, this may be attributed to an in vitro phenomenon of the glucocorticoids and the iron chelators on apoptosis. In this study thymocyte apoptosis by iron chelators could also be detected by microscopy, where the classic features of cell shrinkage and condensation of the chromatin were clearly evident (Figure 3.1b). However the problem with microscopy and agarose gels, as methods for detecting apoptosis, is that they are qualitative and not quantitative, therefore flow cytometric analysis of the sub-diploid peak was used to quantify the amount of apoptosis in cell suspensions. Results in this section showed that the iron chelators, CP20 and DFO increased the amount of apoptosis as measured by flow cytometry (Figures 3.3 & 3.4). This increase in apoptosis as measured by flow cytometry was also qualitatively observed using proportional analysis of CD4^ CD8^. Evidence suggests that CD4'^8^ thymocytes represent the critical stage in T cell development at which the specificity of a randomly generated a/p T cell receptor (TCR) is screened for self reactivity (Kisielow and Von Boehmer, 1990). Experiments using thymus organ cultures treated with anti-CD3 antibodies (Smith et al., 1989) have suggested that death of auto-reactive CD4^8^ thymocytes occurs as a consequence of apoptosis, triggered by TCR-mediated transmembrane signalling. Work presented in this chapter showed that murine CD4^8^ thymocytes expressed lower than normal levels of CD4 and CD8 molecules (Figure 89 3.5), this is in agreement with other investigators who have suggested that these low levels of CD4 and CDS molecules represent a sub-population of immature thymocytes undergoing apoptosis. This idea is supported by the observation of almost complete fragmentation of the DNA isolated from this population (Swat et a i, 1991). Furthermore, Chow et a i, (1997) suggested that the susceptibility of thymocytes to apoptosis is developmentally regulated and showed that double positive thymocytes (CD4+CD8+) and their precursor cells (CD4"CD8+CD3") were equally sensitive to apoptosis after treatment with apoptotic stimuli. They concluded that there was a reduction in thymocyte sensitivity to apoptosis which occurred after the onset of CD8 expression. Lesage et a l, (1997) suggested that the CD8^ and CD4^ thymocytes are preferentially induced to die following CD45 cross-linking and showed that mature thymocytes were much less sensitive to CD45 cross-linking than CD4^ CD8^ cells. The authors suggest that during T cell development, CD45 ligation could induce the death of those immature thymocytes that do not fulfill the requirements for positive selection. The differences observed in the rates of apoptotic induction by CP20 and DFO may have clinical relevance.Figure 3.6 shows that there was little difference in the degree of apoptosis between the chelators following 24h continuous incubation at low chelator concentrations, but at a concentration above 100|iM IBE, CP20 induced apoptosis to a greater extent than DFO, p<0.05, as measured by the students T test. A single oral administration of CP20 at clinically relevant doses results in maximum plasma concentrations of 100-300)liM, whereas intra-venously or sub-cutaneous infusions of DFO at standard doses (50mg/kg/24h) results in plasma levels significantly below this value of between 5 and 20|iM (Lee et a l, 1993). This difference in chelators may be explained by the lipophilic nature of CP20 compared to DFO. Lipophilicity as measured by the partition coefficient (kpart) between octanol and water, influence the uptake and removal of iron. Kpart values close to 1 signify approximately equal solubility in each phase. It has been established that the optimal balance between toxicity and efficacy of an iron chelator is attained when the kpart is between 0.2 and 1.0 (Porter et aL, 1988) The hydrophilic properties of DFO, kpart 0.01, positive charge (4-1), (although forms a neutral complex with iron) and high molecular weight (657molwt) prevents it from penetrating most cells. CP20, kpart 0.2, on the other hand, by virtue of its lipophilicity, relatively small molecular weight (175molwt) and neutral charge in its free and complexed state (pKa of the free ligand is close to 9.0) would be able to penetrate cellular membranes and organelles at a faster rate than DFO, thereby exerting its toxic effects. Futhermore as suggested in section 1.5, CP20 with its high binding constant for ferric iron is able to mobilise iron from ferritin in vitro significantly more rapidly than larger ligands such as DFO (Brady et al., 1988). Indeed results in Table 3.1 showed that although there was very little difference in the amount of apoptosis between the chelators at time points up to 8 hours. 90 by 24h CP20 had a significant increase in the amount of apoptosis compared to DFO. This may be explained by the ability of CP20 to gain access across the nuclear membrane at a faster rate than DFO in order to exert its apoptotic effect. It is possible that this difference in the apoptotic induction rate may contribute to the greater propensity of CP20 compared with DFO to cause thymic atrophy and bone marrow hypoplasia in laboratory animals and agranulocytosis in patients. 3.3 Apoptotic induction by iron chelators in CD34+ ceils in cu ltu re. 3.3.1 Rationale The antiproliferative effects of iron chelators on haemopoietic progenitors in colony forming assays have been described (Porter et al., 1989; Porter et at., 1991) Hoyes et at., 1993) but the effects on apoptosis are unknown. The potential advantage of using CD34^ cells isolated from human peripheral blood to study the effects of iron chelators on haemopoiesis is that any effects induced by iron chelators can be related to the proliferative state of the cells (Go when first isolated) and to their state of differentiation. Culture of CD34^ cells under defined conditions (see below) can result in proliferation and differentiation of these cells (stem cells and progenitors) into later progenitors then into recognisable precursor cells of defined lineages and finally to mature cells of the myeloid lineage (see Figure 3.7). By exposing the cultured cells to iron chelators at a range of time points after initiation of the culture, the susceptibility of cells at different stages of proliferative activity and differentiation could therefore be examined. Haemopoiesis is the term used to describe the formation and development of blood cells (red cells, white cells and platelets). Cellular differentiation, proliferation and maturation of blood cells takes place in haemopoietic tissue, which in a human adult consists mainly (>95%) of bone marrow, but may also include the liver, thymus, spleen and lymph nodes in foetal life. Most blood cells are short lived and new cells must be continuously produced to maintain appropriate amounts of each. The bone marrow has to adapt to the variable requirements for new blood cells and to increase the production rate of all specific cell lineages as necessary (Lowenberg and Touw, 1993). Therefore the life and death of haemopoietic blood cells is a tightly regulated process controlled by membrane receptors and their ligands which can activate proliferative or apoptotic responses to regulate normal homeostasis in haemopoietic blood cells. Current thinking suggests that the blood cells originate from a self-renewing population of multipotential haemopoietic stem cells located in the bone marrow. This monophyletic theory allows haemopoietic cells to be divided into three cellular compartments: 1. the primitive multipotential cell capable of self-renewal and 91 Pluripotent stem cell Lymphoid stem cell (O ro ICFU Thym us CFU-M CFU-G Red Platelets Mono- Neutro- Eosino- Baso- Lymphocytes cells cytes phils phils phils Figure 3.7: Haemopoiesis. Diagrammatic representation of the bone marrow pluripotent stem cell and the lines that arise from it. Various progenitor cells can now be identified by culture in semi-solid medium by the type of colony they form. BFUE, burst-forming unit, erythroid; CPU, colony-forming unit; E, erythroid; Eo, eosinophil; GEMM, mixed granulocyte, erythroid, monocyte, megakaryocyte; GM, granulocyte, monocyte; Meg, megakaryocyte. Taken from Essential Haematology, A.V.Hofthrand and J.E.Pettit. differentiation into all blood lines and 2. which generates proliferating progenitor cells committed to either erythropoiesis, myelopoiesis or thrombopoiesis and 3. mature cells with specialised functions that have lost the capacity to proliferate(Figure 3.7). The more mature cells can be identified by their morphology as they differentiate into mature cells of a particular lineage. They can also be identified by virtue of functional capability and the presence or absence of a number of marker molecules from the CD series (cluster of differentiation). Although a cell surface marker unique to stem cells has yet to be discovered, there are markers that are shared by stem cells and more committed progenitor cells eg. CD34. An important antigen which has been shown to be expressed on haemopoietic progenitors was identified by Civin et a l, (1984) and is now termed the CD34 antigen (formerly known as MYIO). This antigen was subsequently shown to be a surface glycophosphoprotein expressed on developmentally early lympho-haemopoietic stem and progenitor cells, small vessel endothelial cells and embryonic fibroblasts. It is present on 1-3% of bone marrow mononuclear cells but also on more primitive haemopoietic progenitor cells (Smelandet a l, 1992; Rusten et al., 1994). Recent data have shown that enriched CD34^ marrow cells can reconstitute haemopoiesisin vivo in humans and primates (Berenson et at., 1991), however the function of the CD34 molecule in haemopoiesis has yet to be clearly defined. The CD34 antigen is detected in vitro on primitive progenitors such as Long-Term Culture Initiating Cells (LTC-IC) and High Proliferation Potential Colony Forming Cells (HPP-CFC) (Civin, 1984). The CD34^ population is heterogeneous and pluripotent stem cells constitute only a minor fraction of the whole CD34^ population(Figure 3.8). Therefore several attempts have been made to further enrich for the most immature human haemopoietic cells by determination of subsets of CD34+ cells (Figure 3.8). In particular available data suggest that the most primitive cells are enriched in the CD34+CD38", CD34+HLA-DR" or CD34+Thy-1 subpopulations (Srour et a i, 1993; Mayani et at., 1994; Humeau et a l, 1996). Both the CD34+CD38" and CD34+HLA-DR" cells represent only a few percent of the CD34+ population and have been found to be enriched in the most immature cells capable of in vitro growth, and give rise to multi-lineage reconstitution in in vivo models of human haematopoiesis in chimaeric animals (Srour et a l, 1993). The stem cell pool comprises approximately 1 to 2x10^ cells, which are responsible for the generation of more than 10^1 cells per day on a continual basis. Only a small number of stem cells are dividing at any one time (5%), most being in the Go/quiescent phase of the cell cycle (Beradi et a l, 1995). The most ancestral stem cells (termed re-populating cells) are capable of extensive renewal and are defined either by their capacity to re-populate haemopoietic and lymphoid organs on a long term basis in irradiated animals (Till and McCulloch, 1961) or by their capacity to sustain long-term 93 Figure 3.8 Stem cell /progenitor cell hierarchy, clonogenic assays and differentiation antigens associated with haemopoietic populations Adapted from Watts, 1999 Increasing cell maturity A. Total haemopoietic cells in adult system (Moore 1996) 10®-1Q7 5x109 5x109 Stem cell Multipotent Lineage Morphologically Mature marrow restricted recognisable blood and tissue' differentiating cells LTRC I------1 CAFC / LTCIC/ PAÔ HPP-CFC / BL-CFC/ CFU-S I------1 CFU-MIx I------1 BFU-E Key to clonogenic assay abbreviations; I------1 LTRC = Long Term Repopulating Cell (in vivo B. Clonogenic assay) ; CAFC = Cobblestone Area Forming GM-CFC Cell, LTCIC = Long Term Culture Initiating Cell; Assays PAô ("delta assay") = Plastic Adherent cells giving rise to CFC; HPP-CFC = High GM-Clusters Proliferation Potential-Colony Forming Cell; BL- I 1 CFC = Blast cell - Colony Forming Cell; CFU-S = Spleen-Colony Forming Unit (in vivo assay); CFU-E CFU-mix = Colony Forming Unit - mixed h lineage; BFU-E = Burst Forming Unit-Erythroid; C. Antigen staining of Givl-CFC = Granulocyte / l^onocyte-Colony Forming Cells; GM-clusters = Granulocyte / CD344- cells Monocyte clusters; CFU-E -i- Colony Forming Units- Erythroid CD34(MY10) CDw90 (Thy-1) CD38 (T10) HLA-DR CD33 (MY9) CD45RA 94 haemopoeisis in vitro. Stem cells must be capable of balanced self-renewal and differentiation, which may occur in one of two ways. First a stem cell may divide into two daughter cells, one of which is committed to differentiate while the other remains in the stem cell pool. Secondly, for every stem cell that produces two daughter cells, both of which differentiate, another stem cell divides to produce two daughter cells that remain in the stem cell pool (Emerson, 1995). The next stage in the differentiation of haemopoietic cells are the progenitor cells. Progenitor cells are the progeny of multipotent stem cell differentiation and possess limited further differentiation potential; they are known as lineage-restricted or committed progenitors in recognition of this characteristic (Metcalf, 1977, 1988; Testa et al., 1985). Progenitor cells of different lineages are not immediately recognisable morphologically and were originally identified by their ability to form colonies in cell culture using semi-solid medium (agar) and hence given the name colony forming units (CPU). Such cells are named depending on the type of mature cell which results from a colony forming unit, e.g. CFU-GM (colony forming unit-granulocyte, macrophage). Progenitor cells are thus a transient cell population formed from stem cells and are characterised by their capacity to form colonies in tissue culture in response to stimulation by haemopoietic growth factors. It is not possible to indicate how many progenitor cells an individual stem cell can generate. This is partly determined by the stimuli applied to the stem cell and how much self-generation results. The final stage of haemopoiesis encompasses approximately 95% of the cells. It represents the proliferative amplification of the differentiated cells as they mature to become fully functional blood cells. The lymphoid lineages consist of the T cell and B cell lineages. The myeloid lineages include the erythroid, granulocytic, monocytic and megakaryocytic lineages. These cells are recognizable and classified by their classical morphological characteristics (Wright and Lord, 1992) (Figure 3.7). Haemopoietic growth factors are indispensible in haemopoietic cultures but the optimal combination of growth factors required is not clear. One reason for this is that most stem cells and progenitor cells simultaneously co-express receptors for more than one growth factor, resulting in synergistic and antagonistic interactions (Metcalf and Nicola, 1992; Tiwari, 1998). Two haemopoietic growth factors that are important for survival and proliferation are interleukin 3 (IL3) and stem cell factor (SCF) (Brandtet at., 1994). IL3 and SCF prevent apoptosis of early and committed progenitor cells. IL6 acts in concert with other growth factors to induce cycling of additional stem and progenitor cells (Leary et al., 1988). Therefore, SCF, IL3 and IL6 are a common growth factor combination for broad expansion across multiple haemopoietic lineages (Williams et al., 1997). As indicated earlier, most stem cells are quiescent (G q/G i), but the cells have an extensive potential capacity for self-renewal and for the production of a large number of 95 committed progenitor cells (Reems and Torok-Stord, 1995; Beradi et al., 1995; Tiwari, 1998). The cell cycle profile of committed progenitor cells has been studied using semi solid culture methods (colony formation). Committed progenitors from human bone marrow as previously outlined are identified as colony-forming units (CPU) or burst- forming units (BFU) such as CFU-GM (granulocyte-macrophage precursors) and BFU- E (erythroid precursors) in methylcellulose culture. Investigations indicate that 30-50% of the CFU-GM and BFU-E are in the S phase of the cell cycle (Fauser and Messnar, 1979; Aglietta et at., 1989). In contrast, approximately 20-30% of myeloid precursors (myeloblasts, promyelocytes and myelocytes) are in S phase of the cell cycle (Aglietta et al., 1989). This indicates that differentiation of committed progenitor cells to morphologically identifiable precursors is associated with a decrease in the proportion of actively dividing cells. Cycling ability further decreases during the late stages of differentiation, and when the terminal elements (red blood cells, mature granulocytes, monocytes, lymphocytes and platelets) are ultimately released into the blood stream they are completely (more than 95%) arrested in the Gq/Gi phase of the cell cycle (Metcalf and Nicola, 1995; Tiwari, 1998). 3.3.2 Isolation and culture of CD34 cells Peripheral blood CD34+ cells were obtained from patients following ESHAP chemotherapy and G-CSF (see section 2.3.7). The CD34+ human haemopoietic progenitors were purified using the Cellpro CEPRATE© SC CD34'*" purification column (Watts et al., 1997). The cells were cultured at a density of 2x10^ cells/ml in culture medium (RPMI containing L-glutamine and supplemented with 10% FCS) together with the growth factors, lOng/ml each of stem cell factor (SCF), interleukin-3 (IL3) and interleukin-6 (IL6) and incubated with either PBS, CP20 or DFO at 300|iM IBE unless otherwise stated for 7 days. Aliquots of 1ml were removed on days 1, 3, 5, or 7 for flow cytometric analysis. Alternatively, cells were cultured as indicated above with the relevant growth factors for 1, 3, 5, 7, 9, 11 or 14 days, after culture on the respective days the cells were incubated for a further 24h with PBS, CP20 or DFO before morphological or flow cytometric analysis (section 2.4.1). 3.3.3 Changes in CD34 cells under conditions of culture; proliferation, differentiation and cell cycle profile 3.3.3.1 Changes in morphology under conditions of culture Morphological analysis of apoptosis was assessed using hght microscopic examination of May Grunwald-Giemsa cytospin preparations(see section 2.3.4). Figure 3.9a shows the typical morphological appearance of freshly isolated CD34^ progenitor cells. The cells were relatively undifferentiated and had a high nuclear 96 A. * # « I » % » Figure 3.9: Effect of culture on morphology of CD34+ cells CD34+ cells were either freshly isolated (A) or cultured for 3 (B), 5 (C), 8 (D) or 14 days (E) with the growth factors IL3, IL6 and SCF. 97 D. E. # 98 to cytoplasmic ratio, the cytoplasm appeared basophilic and there was an open nuclear pattern. After 3 and 5 days in culture with the growth factors SCF, IL3 and IL6, the cells had developed features of proliferation with several mitotic nuclei present(F ig u re 3.9b & c). After 8 days in cultuie, the cells showed features of myelo-monocytic differentiation, including the formation of myeloid granules, nuclear indentation and in some instances segmentation (Figure 3.9c) The cells also showed monocytic chaiacteristics which included being large and vacuolated with a ‘ground glass’ cytoplasm. By 14 days in culture the majority of cells appeared to be neutrophilic (Figure 3.9d) with condensed nuclear chromatin, nuclear segmentation and the presence of granules. These results were further confirmed by cytochemical and surface marker analysis as described below. 3.3.3.2 Changes in cell number under conditions of culture Freshly isolated CD34^ cells were set up in culture with IL3, IL6 and SCF (section 3.3.2) at a density of 2x10^ cells/ml in RPMI-1640 medium. The cell number was counted on days 3, 5, 7, 9, 11 and 14 of CD34^ differentiation. Cell number increased exponentially until day 11 (Figure 3.10) and showed a doubling time of 24h. After day 11 cells no longer proliferated. Pluripotent stem cells proliferate during the early stages of differentiation in response to growth factors. They gradually lose their proliferative capacity with differentiation and arrest in theG q/G i phase of the cell cycle as functionally mature cells (Metcalf, 1989). 3.3.3.3 Changes in surface antigen expression under conditions of c u ltu re The use of specific combinations of growth factors offers the potential of selective generation or expansion of particular haemopoietic populations(se ctio n 3 .3 .1 ). Using monoclonal antibodies (section 2.6) to specific surface antigens expressed on haemopoietic populations (CD34, CD33, CD 13, CD19 and CD7) it was possible to define differentiated cell populations based on certain phenotypic characteristics (section 3.3.1) Both stem cells and committed progenitors express the CD34 antigen, but the former lack or have very low levels of antigens associated with lineage-specific differentiation (Watt and Yasser, 1992). It was the purpose of this experiment to differentiate CD34’*’ cells using the specific growth factors as outlined in section 3.3.2, followed by flow cytometric analysis to determine the percentage of cells staining positive for specific surface antigens. Figure 3.11 shows the percentages of either 0034"^ (stem cell), CD33^ (early myeloid), CD 13^ (late myeloid), CD 19^ (B cell) and CD7^ (T cell) on days 3, 5, 8, 11 and 14 of CD34 differentiation with 1L3, 1L6 and SCF. On day 3, 87% of the cells stained positively for CD34, 20% were CD33^ and the cells were negative for CD7, 99 SCF+ IL3+ IL6 0) i £ 5 T5 S Day 0 Day 2 Day 3 Day 5 Day 7 Day 9 Day 1 1 Day! 4 F igure 3.10: Effect of culture on CD34^ cells on cell number CD34^ cells (2x10^ cells/ml) were incubated with SCF, IL3 and IL6 shortly after isolation. Aliquots of 1ml were removed on days 2, 3, 5 ,1 ,9 , 11 or 14 and cell number counted. 100 SCF+ IL3+1L6 100 c 8 0 - o s s a . X LU 6 0 - C CD34 O)Q) 2 CD! 3 c CD33 < 0) 4 0 - I3 œ Q) u 20 - * Day 0 Day 3 Day 5 Day 8 Day 1 1 Day! 4 Figure 3.11: Effect of culture on surface antigen expression of CD34^ cells CD34^ cells (2x10^ cells/ml) were incubated with SCF, IL3 and IL6 shortly after isolation. Aliquots of 1ml were removed on days 3, 5, 8, 11 and 14 and using monoclonal antibodies to CD34, CD 13 and CD33, the percentage of these surface antigens were assessed by flow cytometry. The data shown are the mean of one experiment done in triplicate. 101 CD 10 and CD 19 indicating that the ceils were myeloid lineage committed. By day 5, 58% of the cells were still CD34^, 36% were CD33^ and 55% CD 13^ indicating that the cells are maturing down the myeloid lineage. By day 8, only 7.5% of the cells were CD34^, the proportion expressing CDl 3^ had increased to 87% and the proportion of cells expressing the CD33 antigen had also increased to 75%. These results indicated that the cells were differentiating down the myelo-monocytic lineage. By day 11, only 2.5% of cells stained positive with the CD34 antibody. The number of cells staining positive with the CD 13 antibody had increased to 77% and the number of cells staining for CD33 had decreased to 32%. Finally by day 14, cells were negative for CD34, 90% of cells were CD 13^ and 32% of the cells were CD33^ indicating that the majority of the cells were well differentiated down the myelo-monocytic lineage. 3.33.4 Changes in cell cycle status under conditions of culture A combination of SCF, IL3 and 1L6 promotes the rapid entry and progression of quiescent CD34^ cells to progress through the cell cycle. Freshly isolated cells showed no evidence of apoptosis or proliferation and all the cells were in theG q/G i phase of the cell cycle. After 3 days in culture with growth factors, the cells had started to proliferate and the profile showed cells with increasing DNA content (Figure 3.12). Exit from G q and entry into G] is associated with an increase in cell size (Darynkiewicz et cd., 1980) which can be analysed using a two colour flow cytometry of DNA stained with PÏ and total cell protein stained with unconjugated FITC. Recently W illiams et al., (1997) and Tiwari, (1998) showed that CD34^ stimulated with a combinastion of SCF, IL3 and 1L6 increased their total cell protein content within 6h of stimulation , signalling entry into G]. By 12h the numbers of cells in G% had increased and at 24h the cells were predominantly in late G|. Tiwari, (1998) also showed that entry into S phase was first detected at 12h and was maximal at 36h. By 48h up to 70% of the cells were in the cell cycle. These data would indicate that the transition from G q to Gi phase, as monitored by increased protein content, occurs around 6 to 12h after stimulation and entry into S phase was observed around 12-24h after stimulation. 3.3.4 Effect on apoptosis of continuous exposure to chelators as measured by quantitative flow cytometry CD34^ cells were isolated and cultured with the growth factors, SCF, IL3 and IL6 under the conditions described in section 3.3.2, Two hours later, either control (PBS) or the chelators,CP20 or DFO, at a final concentration of 300|lM IBE were added into the CD34"^ suspension. Aliquots of 1ml were removed for flow cytometric analysis of the sub-diploid peak on days 1,3, 5 or 7 as described in section 2.4.1. 102 Day 3 CD34 Cells (SCF, IL3, IL6) G2/M S-phase Apoptosis Increasing DNA content Figure 3.12; Typical flow cytometric profile of proliferating CD34^ cells CD34^ are set up in culture with the growth factors, IL3, IL6 and SCF (lOng/ml). After 3 days a sample containing 1x10^ cells/ml was removed, fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. Increasing DNA content is shown on the x-axis against the number of cells on the y-axis. Figure 3.13a shows that on day 1 the majority of cells were in the G q/G i phase of the cell cycle and only a small amount of apoptosis was evident (7.1±1.2%, 8.4+0.9% and 8.9+0.7% for PBS, CP20 and DFO respectively). By day 3 there was a significant amount of apoptosis in cells cultured in the presence of CP20 or DFO, causing 49.6+5.3% and 40.7+2.9% apoptosis respectively compared to control, 7.0+1.3%, p<0.001. By days 5 and 7 of continual incubation the amount of apoptosis was further increased in chelator-treated CD34^ cells compared to control. 3.3.5 Effect of duration in culture prior to incubation with chelators on apoptosis after 24h exposure. CD34^ cells were cultured under the conditions described in section 3.3.4 for between 1 and 14 days. On days 1, 3, 5, 7, 9, 11 or 14 either PBS, CP20 or DFO (300|liM IBE) were added to cells at 2x10^ cells/ml for a further 24h incubation at 37‘^C/5%C02 and subsequently examined for apoptosis by flow cytometric As shown in Figure 3.13a, in Figure 3.13b at day 1, all cells were quiescent in the G q/G i phase of the cell cycle and very little apoptosis was evident (5.2±0.9%, 8.1+1.3% for PBS and CP20 respectively). By day 3, (by which time the cells had started to proliferate) (section 3.3.3.4), using flow cytometric analysis, there was a significant increase in the amount of apoptosis in chelator-treated cells with CP20 and DFO causing 21.9+2.6% and 21.7+2.6% apoptosis respectively, compared to control, 8.9±1.7% apoptosis. A maximal increase in the amount of apoptosis between chelator treated and control cells was seen on day 5. By day 7 and 9 the amount of apoptosis in chelator-treated cells had decreased with CP20 and DFO causing 15.3+1.4% and 17.4+3.1% apoptosis respectively at day 9, a significant decrease compared to the amount of apoptosis seen on day 7, 22.3±2.1% and 23.3±3.0% for CP20 and DFO, p<0.05. 3.3.6 Effect of chelator concentration on apoptosis following 24h exposure at various time points after initiation of CD34+ culture. Under the same conditions as described in section 3.3.5, CD34-I- cells were cultured for between 1 and 9 days. Aliquots approximately 2xl0^/ml cells were removed on days (1, 3, 5, 7 or 9) and incubated for a further 24h with either PBS or varying concentrations of chelators 0, 11, 33, 100 or 300pM IBE at 37^C/5%C02 prior to analysis by flow cytometry. There was no significant difference between control and the various concentrations of either CP20 or DFO on day 1, all the cells were in the G q/G i phase of the cell cycle and only a small amount of apoptosis was evident(Table 3.2). By day 5, CP20 at concentrations of 300, 100 and 33pM IBE was found to cause a significant amount of apoptosis (24.3+0.7%, 24.2+0.6% and 17.3+0.4 respectively) compared to 104 Addition of drugs SCF+IL3+IL6 100 8 0 - PBS CO •3! o 6 0 - CP20 o a. < » DFO 4 0 - 20 - "I' " I ...... ' ■" '" 'p " DAY 1 DAY 3 DAY 5 DAY 7 Figure 3.13a: Apoptosis in CD34^ cells with continuous exposure to iron chelators CD34+ cells (2x10^ cells/ml) were incubated with SCF, IL3 and IL6 and either PBS, or chelators, 300pM IBE for 24h (1), 72h (3), 168h (5) or 216h (7). The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ±SD of 4 independent experiments done in triplicate. 105 SCF+IL3+IL6 30- I n w « 2 0 - o a. ■ PBS o 0 CP20 a. < □ DFO i DAY 1 DAY 3 DAY 5 DAY 7 DAY 9 DAY 1 1 DAY 1 4 Figure 3.13b: Effect of timing of exposure of CD34+ cells to chelators on apoptosis CD34^ cells (2x10^ cells/ml) were incubated with SCF, IL3 and IL6 shortly after isolation, then aliquots were removed on days 1,3,5, 7, 9, II or 14 and either PBS, or the chelators, 300pM IBE, added for a further 24h incubation. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ±SD of 4 independent experiments done in duplicate. 106 Sample Day 3 % APOPTOSIS PBS 5.2±0.5 10.0±0.8 11.3±0.4 , 300pM IBE 7.4±1.0 20.8±2.4** 24.3±0.7** lOOpM 7.8±1.3 23.0±2.8** 24.4±0.6** 33 pM 5.9±0.4 17.7±1.3* 17.4±1.1* llpM 5.1 ±0.9 17.4±1.1* 17.4±1.1* 300pM IBE 8.2±1.0 22.2±0.1** 25.6±3.2** lOOpM 7.3±1.6 22.8±1.7** 20.1±1.2** 33pM 6.1 ±0.3 14.4±0.1* 13.6±0.2 llp M 6.7±1.4 14.9±1.8* 13.3±0.4 Table 3.2; Effect of chelator concentration on CD34^ cell apoptosis CD34'*’ cells were incubated with SCF, IL3 and IL6 shortly after isolation. On day 1, 3 and 5 aliquots were removed and incubated with either PBS or chelators at 11, 33, 100 or 300[xM IBE for 24h. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. * and ** denotes significance of p<0.01 and p<0.05 respectively as measured by the students T test. 107 control (11.3±0.4%). DFO, only at concentrations of 300 and 100|liM IBE was found to significantly increase the amount of apoptosis (25.6±3.2% and 20.1+1.2%) compared to control. 3.3.7 Pulsed chase effect of 6h exposure to chelators: Subsequent effect on apoptosis In order to determine whether a short exposure of the cells to chelators was sufficient to induce apoptosis immediately or later, CD34^ cells were cultured as in section 3.3.5 except that after 12h in culture when the cells were quiescent in Gq/G], they were pulsed for 6h with either PBS or the iron chelators, CP20 or DFO (300pM IBE) at 37^C/5% CO 2 . After this time, the cells were washed 3x and resuspended in culture medium with the relevant growth factors. After 12h in culture with the growth factors, the cells were washed a further 2x and resuspended in culture medium with growth factors in order to be sure that any residual chelator had been sufficiently washed out. Aliquots were then removed on days 2, 3, 5 and 7 and analysed by flow cytometry. Alternatively CD34^ cells were cultured for either 5 days or 14 days and exposed to a short pulse (6h) of chelators to examine both the effect of a short pulse of iron chelators when the cells are actively proliferating oi' on cells which have become mature. Once again, after a 6h incubation with chelators, the cells were washed 3x with a further 2x washes after a short 12h incubation as indicated above and resuspended in culture medium with the growth factors. For experiments to investigate the effects of a short pulse on proliferating cells at day 5, aliquots of 1ml containing 2x10^ cells were removed on days 7, 9 and 11 and analysed by flow cytometry. To investigate the effects of iron chelators on mature cell types at day 14, aliquots of 1ml containing 2x10^ cells were removed on days 15 and 17. Figure 3.14a shows the effect of a short pulse, 6h, of iron chelators on CD34^ apoptosis when the cells are in the Gq/Gj phase of the cell cycle. Only a small amount of apoptosis was evident on day 2 (8.1+0.9%, 8.7+1.4% and 9.5+1.4% for PBS, CP20 and DFO respectively). By day 3 when the cells were cycling (section 3.3.3.4) no increase in apoptosis compared to control was evident. With samples analysed on days 5 and 7, there was still no evidence of an increase in apoptosis compared to control despite the proliferative status of the cells. To investigate this further, proliferating cells, either early progenitors or mature were exposed to a short pulse of chelators. Figure 3.14b shows the effect of a short pulse (6h) of iron chelators on cycling CD34+ cells in culture with growth factois for 5 days. After 5 days in culture CD34+ cells were myeloid lineage comiuitted and had begun to differentiate (section 3.3.3.3). By day 7 there was a small amount of apoptosis in control cells (10.2+2.3% ), however in comparison to the results presented inFigure 3.14a, both 108 Addition of drugs for 6h SCF+ IL3 + IL6 20 15 v> -D '55 PBS o ^----- CP20 o. o *=----- DFO a. 10 < * 5 0 Day 2 Day 3 Day 5 Day 7 Figure 3.14a: Effect of a 6h pulse of chelators on day 1 of CD34^ culture Figure 3.14: Effect of pulse of chelators on differentiating CD34^ cells CD34^ cells (2x10^ cells/ml) were incubated with SCF, IL3 and IL6 shortly after isolation. A 6h pulse of chelators, 300pM IBE was added on either day 1 (Fig.3.14a), day 5 (Fig. 3.14b) or day 14 (Fig. 3.14c). The cells were washed, fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean of 2 independent experiments done in triplicate. 109 Addition of drugs for 6h 1L3+ 1L6 40 ■ 35 ■ 30 - U) (A 25 • O 4-1 a . o Q. 20 ■ PBS < CP20 DFO IS ■ 10 - 5 ■ 0 ■ Day 5 Day 7 Day 9 Figure 3 .14b: Effect of a 6h pulse on day 5 of CD34^ culture Addition of drugs for 6h 20 PBS 10 - CP20 DFO 5 - Day 14 Day 1 5 Day 17 Figure 3.14c: Effect of 6h pulse of chelators on day 14 of CD34^ culture 110 the chelators caused a significant amount of apoptosis (22.3+3.2% and 24.1+2.9% for CP20 and DFO respectively) which had increased by day 9 (29.1±1.8% and 31.7+3.3% for CP20 and DFO respectively compared with control 1 1.5+3.8%, p<0.05). These results would suggest that the effect of the chelators on apoptosis of CD34^ is cell cycle dependent, and this will be addressed further in chapter 4. Figure 3.14c shows the effect of a short pulse (6h) of iron chelators on CD34+ cells in culture with growth factors for 14 days. Results from section 3.4.3 3 indicated these cells to be relatively mature and were no longer proliferating (section 3,4.3.2). When cells were analysed on days 15 and 17, no significant increase in the amount of apoptosis was seen in chelator treated cells from control (13.5+1.9% for PBS on day 15 compared to 15.1+1.7% and 13.8+3.1% for CP20 and DFO respectively). With regard to propensity to apoptosis, these results would suggest that the chelators are unlikely to affect mature cells. 3.3.8 Discussion. In the microenvironment of the bone marrow, stem cells are subjected to a range of stimuli including physical interactions with other cells and exposure to both growth stimulatory and growth inhibitory cytokines. Together these stimuli control self renewal and differentiation of stem cells and their subsequent evolution into mature cells in the blood. Much work has concentrated on growth factor withdrawal resulting in stem cell apoptosis and death (Philpott et ai, 1997). In the present study CD34+ human haemopoietic progenitor cells were used as a model for the bone marrow. CD34+ cells are quiescent when isolated but can be induced to proliferate, therefore making it possible to investigate at what stage of proliferation cells were being affected by the iron chelators and the question would be answered using the one cell type. CD34^ cells have been shown to be in a quiescent state when isolated (Williams et al, 1997; Tiwari, 1998). These cells have a 2n DNA content and are small with low total protein compared with their proliferating counterparts. Tiwari, (1998) using a two colour flow cytometry of DNA and total cell protein showed that after stimulation with a combination of growth factors (IL3, IL6 and SCF) there was an increase in total cell protein after 6h indicating cellular entry into G ]. The author suggested that transition from Go to G t phase, as monitored by increased protein content, eommenced around 6 to 12h after stimulation, with entry into S phase observed around 12-24h after stimulation. Data presented in this chapter suggest that iron deprivation by the addition of the iron chelators, CP20 and DFO can cause a significant increase in programmed cell death in early proliferating progenitor cells as assessed using morphological and flow cytometric analysis (Figures 3.9, 3.13). Quiescent cells were found to be unaffected with respect to the amount of apoptosis by iron chelators (Figures 3.13, 3.14a), 111 however by day 3 of growth factor stimulation there was a significant increase in the amount of apoptosis compared to control(Figures 3.13a & b, Table 3.2). As suggested by Tiwari, (1998) this increase in the amount of apoptosis could be attributed to the cell cycle status of the cell and may serve as a rationale for further studies of iron deprivation as a form of cancer treatment. After 3 days in culture with growth factors, CD34^ cells had begun to differentiate down the myelo-monocytic lineage as determined by microscopy and surface antigen analysis. Hoyes et al., (1993) showed that in vitro micromolar concentrations of both DFO and CP20 wei e able to inhibit colony formation (CFU-G and BFU-E) in a dose dependent manner. Indeed the authors showed that the addition of iron to saturate the chelators abrogated the effects of DFO, but not of CP20. With the CP20-iron complex, only the addition of sufficient iron to saturate the transferrin in the medium reversed the inhibitory effects. As suggested by the authors, CP20-iron complexes may dissociate both in the extracellular medium with donation of iron to transferrin, and intracellularly with the released iron being sequestered into intracellular pools whereby the free CP20 would be available to interfere with any iron dependent process. Studies in this section showed that there was no significant difference between CP20 and DFO at inducing apoptosis in CD34^ cells and their progeny, even at low concentrations of chelator. Of interest is that by day 9 of CD34"^ cell proliferation and differentiation (Figure 3.13b) the amount of apoptosis has significantly decreased compared to that shown on day 7. By day 9 of in vitro differentiation, cell surface antigen staining suggested that the cells are relatively mature myeloid cells (Figure 3.11) showing features of neutrophils (Figure 3.9). Furthermore, although the amount of apoptosis was reduced at day 9 as compared to day 7 the cells were still dividing as determined by cell number (Figure 3.10), but not to the extent that was occurring on days 3-7. The significant decrease in apoptosis may be explained by several hypotheses. As outlined in section 1.3.1, cells require iron for proliferation and maturation. Proliferating cells express high densities of transferrin receptors on the surface of their plasma membrane and take up iron by receptor-mediated endocytosis (Ponka, 1998). Therefore the number of transferrin receptors by day 9 may be reduced as compared to that on days 3-7. Indeed by day 14 when the cells had stopped proliferating (Figure 3.10) and showed features of mature neutrophils, very little apoptosis was occurring(Figure 3.13b). These results would indicate that both CP20 and DFO affect an early proliferating population of cells. This conclusion is also substantiated by results from pulse-chase experiments (Figure 3.14a, b & c). Exposing quiescent cells (G q/G i) with chelators for a short time and resuspending the cells in fresh medium supplemented with growth factors, analysis when the cells were in cycle showed no evidence of increased apoptosis (Figure 112 3.14a). In comparison, when the cells were in cycle and exposed to a short pulse of chelators, a significant amount of apoptosis was seen (Figure 3.14b), Indeed when the cells were exposed to a short pulse on day 14, when the majority of the cells were mature neutrophils and were no longer proliferating(sections 3.3.3.1 and 3 .3 .3 .2 ) there was no increase in the amount of apoptosis compared to control suggesting that chelators are unlikely to affect mature cells with regard to apoptosis and the effects of the chelators on CD34^ cells and their progeny are likely to be cell cycle dependent. 3.4 General discussion and conclusions In this section the propensity of iron chelators, CP20 and DFO to induce apoptosis in quiescent thymocytes and quiescent and proliferating human haemopoietic progenitors was examined. Normal development of cells in the bone marrow requires the regulation of cell viability, growth and differentiation to different lineages. Different cell types in the body require specific factors to remain viable (Fibach el al., 1976; Lotem et al., 1989; W illiam s eî al., 1990; Raff, 1992) and removal of such factors results in apoptosis. One important viability factor is iron. It is known that iron is essential for growth and viability and that iron deprivation results in inhibition of the synthesis of deoxyribonucleotides, in particular by inhibiting the iron-containing enzyme ribonucleotide reductase (Hoyes et ai, 1992; Cooper et al., 1996). Furthermore, in this study it was shown that iron deprivation using the iron chelators, CP20 and DFO induced a significant amount of apoptosis in thymocytes and proliferating haemopoietic progenitors. The chelators did not increase the amount of apoptosis compared to control in quiescent CD34^ cells, apoptosis was only increased in proliferating CD34^ eells. It is possible therefore that the effects of the iron chelators on apoptosis in CD34^ cells are either cell cycle dependent or that apoptosis is an effect of cell differentiation. The latter hypothesis however appears unlikely since studies in this section showed that it was the early haemopoietic progenitoi's which were affected by the iron chelators and the more mature and differentiated the cell was, the less susceptible to apoptosis by iron chelators it becam e (Section 3.3.5). In thymocytes a greater degree of apoptosis was seen with CP20 compared to DFO. In thymocytes there was a significant difference in the degree of apoptosis between CP20 and DFO after 24h incubation at 300p.M IBE, with little difference in the amount of apoptosis at low concentrations of chelators, <100p.M IBB. As discussed in this section, such differences between the chelators in thymocytes may be exploited by the chemical nature of CP20 in comparison to DFO. Both CP20 and DFO have distribution coefficients of 0.2 and 0.01 respectively. Hence according to their distribution coefficients, CP20 has a greater ability to penetrate cellular 113 membranes and gain access to intracellular iron pools (compounds with kpart values between 0.1 and 1 show a steady increase in their ability to mobilise intracellular iron) Alternatively the increased amount of apoptosis with CP20 compared to DFO in thymocytes may be explained by the modes of chelation of iron from this compound. DFO cannot chelate iron from transferrin, whilst CP20 can (Evans et al., 1992; Al- Refaie et al., 1995). CP20 has also been shown to release iron from other stores such as ferritin and its iron cores in vitro whereas DFO cannot (Brady et al., 1988) (section 1.4). Such differences may contribute to the main clinical toxicities of CP20 including thymic and other organ atrophy in animals (Hershko, 1993). However, the results in this section do not explain the clinical picture of severe neutropenia and agranulocytosis seen in patients receiving CP20. Proliferating cell types such as CD34^ human haemopoietic progenitors showed little difference in the degree of apoptosis between the chelators at high and low concentrations (300pM IBE and 1 IpM IBE). It may be that differences can only be seen between CP20 and DFO in human progenitors using lower concentrations of the drugs, eg. <10pM IBE. However, although no appreciable difference between the chelators with regard to the amount of apoptosis was seen it may be that the chelators are affecting the cell cycle of CD34^ cells and differences between CP20 and DEO are only obvious in certain phases of the cycle. As suggested in section 3.3.7 the effects of the chelators on apoptosis may be cell cycle dependent. Therefore one can hypothesize that the pattern of toxicities of CP20 and DEO are that of an anti-proliferative compound. Indeed several authors have suggested that interference by the iron chelator on cell proliferation by removing iron, eg. by inhibiting intracellular ribonucleotide reductase is the mechanism of apoptosis (Porter et al., 1994; Hileti et al., 1995; Haq et al., 1995). This would certainly seem a plausible hypothesis in proliferating cell types such as haemopoietic progenitors, with results indicating that it was the proliferating cells which apoptosed. Several independent lines of evidence have fostered the notion that there is a link between cellular proliferation and apoptosis (Evanet al., 1995; King et al., 1995). Key tumour suppessor proteins such as retinoblastoma and p53 exert effects both on cell viability and on cell cycle progression. Furthermore, damage to the cell cycle or to DNA integrity, eg. by affecting many of the important enzymes involved in DNA synthesis may contribute to the onset of apoptosis. Investigation as to whether CP20 and DFO exert their apoptotic effects through a cell cycle mechanism is therefore investigated in the next chapter. 114 CHAPTER 4 The Relationship Between the Anti-proliferative and Apoptotic Actions of Iron Chelators 115 4.1 Introduction In Chapter 3, it was established that the iron chelators CP20 and DFO induce apoptosis in many cell types including murine thymocytes and haemopoietic progenitors. It has long been recognised that iron chelators have anti-proliferative effects (Renton et al, 1996; Alcain et al, 1997; Kyriakou et al., 1998; Juckett et al, 1998) and inhibit DNA synthesis, arresting proliferation in late Gj/S border (Lederman et al, 1984; Wang et al, 1989; Hoyes et al, 1993). There is now also good evidence that the iron containing enzyme, ribonucleotide reductase (RR) is inhibited by iron chelators with kinetics which are compatible with the observed effects on DNA synthesis and cell cycle (Hoyes et al, 1992; Cooper et al, 1996). The inhibition of DNA synthesis secondary to RR inhibition could in principle be a trigger to apoptosis by iron and it is therefore the aim of this chapter to establish whether inhibition of RR is necessary and sufficient for the induction of iron chelator- induced apoptosis. G(/Gi G^/M V % k JS B 3 % > % cs I DNA content Figure 4.1: Hypothetical flow cytometric profile of the cell cycle 116 Iron is an essential element for cell growth and division because many iron- containing proteins catalyse key reactions involving energy metabolism (oxidation and reduction reactions), respiration and DNA synthesis. Without iron, cells are unable to proceed from Gi to S of the cell cycle (Lederman et al, 1984). The cell cycle consists of a series of stages whereby the DNA content of cells in G q/G 1 phase is increased by synthesis in S-phase resulting in twice the content of DNA in M phase immediately prior to mitosis with the resultant formation of two daughter cells (Figure 4.1). Most cells in mammals exist in a non-proliferating state (G q) and progression of such cells from Gq to G, requires stimulation by growth factors. The stimulated cells then carry out a program of commitment, which requires the synthesis and accumulation of new proteins which are present at low concentrations in quiescent cells (G q). The synthesis of a number of specific enzymes markedly increases at the G /S border. Such enzymes include thymidine kinase, ribonucleotide reductase and DNA polymerase-a, all of which are required for DNA synthesis. Hence all cells require iron and indeed neoplastic cells have a high requirement related to their rapid rate of proliferation and this is reflected by an increase in the expression of the transferrin receptor (Galey, 1997). As outlined in Section 1.1, an important iron-containing enzyme is Ribonucleotide Reductase. Ribonucleotide Reductase (RR) is an iron dependent enzyme responsible for the synthesis of deoxyribonucleotides which are required for DNA synthesis (Figure 4.2) and is rate limiting for the transition of cells through Gi. All RR's catalyse the substitution of the OH-group at the position 2' of ribose by a hydrogen, with NADPH as the ultimate hydrogen donor: NADPH + H+ 4- ribonucleotide ------> N A D P+ + deoxyribonucleotide +H 2O Three distinct classes of RR have been described (Thelander et a i, 1979, Vincent et al., 1990, Kayyali, 1997, Fontecave, 1998): Class 1, which has an a2p2 protein structure occurs in all mammalian cells, DNA viruses of the Herpes group and in some prokaryotes in particular, Esherichia coli. This class has an oxygen linked iron centre together with a tyrosyl radical, both of which are essential for enzymatic activity (Reichard, 1993); Class 2 is present in many prokaryotes, with Lactobacillus leichmanni as the prototype. This class requires adenosylcobalamine as a cofactor; Class 3 is present in anaerobically grown E. Coli and in crude extracts of Methanobacterium thermoautotrophicum. It uses S-adenosyl methionine as a cofactor to generate a radical on a glycine residue of the polypeptide chain. 117 The iron centre in this class of the enzyme is an iron-sulphur cluster (Licht et al, 1996; Ollagnier et al, 1996). In spite of these structural dissimilarities, the mechanism of reduction at the substrate level appears to be similar (Que, 1991). PPO 3 Ribonucleotide PPO Reductase R-(SH) 2 R-S 2 + H 2 O OHOH OH Nucleoside Diphosphate dATP Ribonucleotide dADP (jQP Reductase Kinase dGTP ^ dG DP dCTP CDP dCDP ► dUM P DNA UDP dUDP- polymerase Thymidylate I Synthase dTMP- -dTTP V Figure 4.2: Schematic pathway for the synthesis of deoxyribonucleoside 5'-triphosphates for DNA synthesis. B-base, R-(SH)2-thioredoxin The mammalian RR is the best characterised (Fontecave, 1998). The enzyme consists of two non-identical subunits named R1 and R2, each composed of two identical polypeptide chains(Figure 4.3). The large subunit, protein R1 (IVOkDa) binds substrates and allosteric effectors and contains the sulfhydryl groups directly responsible for the reduction of substrate during the enzymatic reaction. The small subunit, protein R2 (SSkDa) contains in each of its two polypeptide chains a pair of antiferromagnetically coupled ferric ions and a tyrosyl free radical essential for enzymatic activity. A 1:1 complex of both subunits is necessary to form an active enzyme. Therefore, because of the redox-active iron at the enzymes active centre it is a target for iron chelators. Indeed the removal of the iron by agents such as DFO has been proposed to underlie the anti-proliferative effects of iron chelators. 118 Substrate Specificity site Control Sites 6 1 Subunit Activity Site SH SH SH SH Catalytic Site 62 Subunit Figure 4.3: Model of Ribonucleoside diphosphate reductase fromE.coli DFO has been shown to have anti-proliferative effects on many cell types such as lymphocytes (Ledermann et al, 1984; Lucas et al, 1995), tumour cells (Beeton et al, 1988; Estrovet al, 1988), peripheral blood mononuclear cells (Kyriakou et al, 1998) and bone marrow (Dezza et al, 1987) and the suggested mechanism for this is inhibition of RR. Although the mechanism of RR inhibition by iron chelators is at present currently unclear: it may involve the iron chelators; a. forming a ternary complex with the enzyme; b. chelating or reducing the iron centre or c. depriving the precursor iron from the newly synthesised enzyme pool (Cooperet al, 1996). Furukaw a et al, 1992 showed that treatment of human leukaemic K562 cells with DFO resulted in a decrease in RR activity, DNA synthesis and cell growth. Indeed exposure of the cells to an anti-transferrin antibody, which blocks iron uptake into cells caused decreased RR activity and inhibited DNA synthesis in a similar fashion as measured by the rate of ['"’C] deoxycytidine formation. Decreases of RR activity and DNA synthesis by the antibody were restored by the addition of ferric nitriloacetate. Such results indicated that RR activity was dependent on the iron-supply and also played a role in the regulation of cell proliferation. Recently Cooperet al, 1996 investigated the kinetics of RR inhibition 119 by the hydroxyj^ridinones (HPO), CP20 and CP94 in comparison to DFO, as measured by electron paramagnetic resonance spectroscopy (EPR). They showed that the rate of decline of the tyrosyl radical in K562 cells incubated with DFO corresponds to the half-life of the enzyme (3-4hrs), suggesting that DFO is inhibiting the incorporation of iron into the newly synthesised enzyme. However, in comparison the HPO chelators removed the radical at a much faster rate than DFO. The authors concluded that the HPO chelators were inhibiting the enzyme by a direct access mechanism allowing them to chelate the iron centre in the enzyme. However, whilst it is clear that iron chelators such as DFO do indeed inhibit RR intracellularly at clinically relevant concentrations (Hoyes et al., 1992; Cooperet a!., 1996), it is not clear that this is the only mechanism of DNA synthesis inhibition or what role this mechanism may play in apoptotic induction. Hydroxyurea (HU) is a well known inhibitor of RR (Zhou et al., 1995; Fontecave, 1998). It is used as a cancer chemotherapeutic drug in the treatment of haematological malignancies such as chronic myeloid leukaemia and essential thrombocythemia and as an inducer of foetal haemoglobin synthesis in haemoglobinopathies. HU inhibits RR by inactivating the tyrosyl radical of the R2 subunit by reducing the ferric iron centre to ferrous iron, which subsequently leaks out of the protein and consequently the enzyme activity is inactivated (Ochiai et al., 1990; Karlsson et al., 1992; Lassmann et al., 1992; N y holm et al., 1993). Whereas, iron chelators as previously mentioned, by limiting the availability of iron, also inhibit the R2 subunit and consequently ribonucleotide reductase activity. In this chapter therefore, the effects of the known RR inhibitor hydroxyurea have been compared with those of the iron chelators CP20 and DFO with respect to induction of apoptosis and inhibition of DNA synthesis in haemopoietic cells in order to determine whether inhibition of DNA synthesis is sufficient or necessary for iron chelator-induced apoptosis. 4.2 Effect of DNA synthesis inhibition on thymocyte apoptosis 4.2.1 Rationale Most thymocytes are known to be depleted from the thymus during T cell development, with the process of thymocyte death considered to be apoptosis.In Vivo thymocytes are predominantly a quiescent (G(/G,) population of cells, with a very small proportion of cells in cell cycle (Cohen 1991).In vitro, proliferation can be stimulated by the addition of growth factors. Having previously shown inchapter 3 that thymocytes apoptosis is increased with the addition of the iron chelators it was the purpose of this section to examine whether inhibition of DNA synthesis was a contributory mechanism of 120 apoptosis in this cell type, by comparing and contrasting the effect of the iron chelators with hydroxyurea, a known inhibitor of the iron containing enzyme RR which is involved in DNA synthesis. 4.2.2 Comparison of the effects of hydroxyurea and chelators in the induction of thymocyte apoptosis using flow cytometric analysis Experimental procedure 1x10^ thymocytes in RPMI-1640 medium were incubated for 24h with either PBS, the iron chelators, CP20 or DFO (300|U,M IBE), HU (ImM) or DEX (10”^M) for 24h at 37®C/5% C O 2, then subsequently treated for apoptosis analysis by flow cytometry as outlined in section 2.4.1 Results. In Figure 4.4a thymocyte apoptosis is shown as quantitated by flow cytometry after 24h incubation with PBS, CP20 (300pM IBE) or HU (ImM). There is a small increase of apoptosis with HU (49.1+3.7%) compared with control (43.2±6.3%) although this fails to reach statistical significance. This contrasts with the large and statistically significant effects with CP20 or DEX on apoptosis (section 3.2). In Figure 4.4b the dose response is shown for the effect of increasing concentrations of HU on apoptosis. A clear dose response effect is seen up to8 mM, although the increase in apoptosis is only small. Discussion. The small increase in apoptosis with HU is consistent with the notion that proliferating cells are susceptible to apoptosis by drugs which inhibit ribonucleotide reductase because only 5.4±0.9% (Section 4.2.3) of freshly isolated thymocytes are in S phase. Indeed the increase in the proportion of thymocytes which undergo apoptosis with HU ( 6 %) is approximately the same as the proportion of thymocytes in S phase. However even at high concentrations of HU (8 mM) the apoptosis is very small compared with CP20 and DEX where the apoptotic induction is considerably in excess of the proportion of cells in S phase. 121 □ PBS □ HU (ft o OQ. a. < CP20 drug used Figure 4.4a: Comparison of apoptotic induction by iron chelators and hydroxyurea Murine thymocytes were incubated with 300pM IBE CP20 or ImM HU for 24h. Control cells received an equal volume of PBS. The cells were washed, fixed in cold 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 122 40 ■ û. a 0 2 4 6 8 10 Drug Cone, (mM) Figure 4.4b: Effect of concentration on hydroxyurea-induced apoptosis Murine thymocytes were incubated with increasing concentrations of hydroxyurea, lOOp-M to 8 mM for 24h. The cells were washed, fixed in cold 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 123 4.2.3 Effect of iron chelators on thymocyte DNA synthesis using 3H- thymidine uptake Rationale Although the majority of thymocytes are not proliferating (quiescent) (Cohen, 1991), there is also a population of proliferating cells. Feamheadet a l, (1994) showed that after incubation of thymocytes in culture, cells that became apoptotic were predominantly from Gq/Gi and G2/M of the cell cycle and not from S phase. The authors concluded that the quiescent (G q/G i), rather than the proliferating cells undergo spontaneous apoptosis in culture. Furthermore, incubation with dexamethasone had no effect on the progression of proliferative cells through the cycle, therefore resembling the pattern of spontaneous apoptosis seen in drug-free cultures. In this section, the aim was to determine whether inhibition of DNA synthesis was necessary for iron chelator induced apoptosis to occur. Experimental procedure Relative DNA synthesis in thymocytes treated with a variety of agents was compared by determining the incorporation of ^H-thymidine into cellular DNA. 1x10^ cells/ml in RPMI 1640 medium were incubated for a short Ih exposure with control, PBS, chelators (BOOpM IBE), hydroxyurea (ImM), dexamethasone (0.1 pM) or aphidocolin (0.1 pM). The samples were then divided into triplicate containing 300,000 cells each in a 200pl volume which were placed in microtitre plates. thymidine (IpCi) was then added to each sample and the incubation was continued for a further Ihour. Following this time, the labelled cells were washed with PBS and collected onto glass fibre filters (GF/C Whatman International Ltd. Maidstone, UK) using an automated cell harvester (Dynatech). Incorporation of ^H-label radioactivity was measured by liquid scintillation counting after the addition of aquasol scintillation fluid, the results are expressed in counts per minute (CPM). Results and Discussion There is a significant inhibition of DNA synthesis (Figure 4.5) after a Ih incubation of thymocytes with HU, p<0.001. CP20 also inhibits DNA synthesis under the same conditions (Figure 4.5), but to a lesser degree than HU, p<0.001, when compared to control and p<0.001 when CP20 and HU are compared. Aphidocolin is a known inhibitor of DNA polymerase and was included as a positive control for inhibition DNA synthesis as previously described (Hughes et a l, 1996). 124 12000 - 10000 - 8000 - S IS c X o CO X ra 6000 I L.I CO O I s 4000 1 2000 - PBS CP20 HU DFO DEX Aphid Figure 4.5: Effect of Ih pulse of chelators and hydroxyurea on DNA synthesis of HL60 cells as measured by thymidine uptake Thymocytes were incubated with either PBS, DEX (lO'^M) the iron chelators CP20 or DFO (300pM IBE), hydroxyurea (ImM) or Aphidocolin (lO'^M) as a positive control for Ih. At the same time the cells were pulsed with l|LiCi thymidine also for Ih. The incorporation of the ^H-label was measured using liquid scintillation counting. The data shown are the mean ± SD of 3 independent experiments done in triplicate. 125 In comparison to CP20, DFO had no significant effect on DNA synthesis after Ih incubation as measured by inhibition of thymidine uptake. As outlined in chapter 3 the ability of chelators to mobilise intracellular iron has been related to their relative lipophilicity, to their charge and their molecular weight. DFO by virtue of its larger molecular weight than CP20 and a Kpart of 0.01 would therefore be predicted to have a slower access to intracellular iron pools and may be one of the reasons for a lack of inhibition of DNA synthesis after Ih incubation. Cooper et al., (1996) showed that DFO took longer to inhibit RR and DNA synthesis in intact cells than HPOs. The authors hypothesised that this might be due to the HPOs, by virtue of their small size, interacting directly with iron within the RR molecule whereas DFO, because of its larger size, could not do this directly and could inhibit RR only by preventing the incorporation of iron into newly synthesised RR. As pointed out above, that the proportion of cells apoptosing after exposure to CP20, far exceeds the number of cells in cell cycle. For example in chapter 3, results showed that after 24h exposure to CP20, 59.8+4.5% of cells had apoptosed(section 3.2), however cell cycle analysis of cells at 24h shows that only 5.4±0.9% are in cycle. These data taken with the proliferation data suggest that CP20 can induce apoptosis in thymocytes which are not proliferating. The findings with HU support this conclusion, HU, while inhibiting the cells in cell cycle at Ih, had no statistical effect on overall apoptosis compared to control(Figure 4.4a). It is hypothesised that only the small proportion of cells in cycle are susceptible to inhibition of RR by HU and the remaining cells which are not proliferating, are not influenced by this mechanism. If this hypothesis is correct, then dexamethasone and iron chelators must be acting by a different (or additional) mechanism(s) of apoptotic induction in thymocytes. 4.3 Effect of iron chelators on proliferating leukaemic cell lines 4.3.1 Rationale In contrast to thymocytes, the majority of cell lines under appropriate culture conditions are rapidly proliferating. Therefore by contrasting the relative effects of HU and chelators on apoptosis and DNA synthesis in cell lines with those of thymocytes, the contribution and importance of cell proliferation to apoptosis can in principle be examined. Advantages of cell lines, compared with primary cell isolates as investigated inChapter 3 are that they are a homogenous group of cells and have the capacity to proliferate continuously in suspension culture. Previously Porteret ah, (1994) showed that iron chelators could induce apoptosis in the K562 erythroleukaemic cell line and Hileti et at.. 126 (1995) showed apoptotic induction using the human myloblastic leukaemia cell line, HL60. Several authors have recently shown that iron deprivation by iron chelators inhibits DNA synthesis and may induce apoptosis (Haq et a i, 1995; Kovar et a i, 1997). Haq et al., (1995) reported iron deficiency and characteristic features of apoptotic cell death with the treatment of the human leukaemic CCRF-CEM cells with DFO, effects which could be prevented by the addition of ferric ammonium citrate. Kovaret at., (1997) showed that the mouse B cell lymphoma cell line, 38C13 underwent apoptosisin vitro when deprived of iron by three independent methods. 1. exposure to a synergistic pair of rat IgG monoclonal antibodies against the mouse transferrin receptor; 2. exposure to the iron chelator, DFO and 3. exposure to a defined culture medium without any added iron. If RR inhibition is the mechanism of apoptotic induction, then apoptosis should be cell cycle dependent and cells should be most susceptible during the S phase of the cell cycle. Additionally, the time and cell cycle dependency of apoptosis on incubation with iron chelators should broadly parallel those of the known RR inhibitor, HU, provided the rates of entry of HU and HPOs into cells are similar. The aims of the work in this section were two fold. Firstly, the relationship between DNA synthesis inhibition and apoptosis has been compared between iron chelators and the known RR inhibitor HU in HL60 cells. Secondly, the relationship between cell cycle and apoptotic induction has been examined using bromodeoxyuridine to label cells in the S phase of the cell cycle. 4.3.2 Effects of hydroxyurea and chelators on apoptosis and DNA synthesis in HL60 cells 4.3.2.1 Effect of chelators and hydroxyurea on apoptosis and cell cycle in HL60 cells using flow cytometric analysis Experimental procedure The leukaemic cell line, HL60 was resuspended at a concentration of 2x10^ cells/ml in RPMI-1640 medium. To investigate the effect of HU on apoptosis, cells were incubated for 24h with either PBS, CP20 or DFO (300|iM IBF) or HU (ImM) at 370C/5% CO 2 and treated for subsequent apoptotic and cell cycle analysis by flow cytometry as outlined in section 2.4.1. Results and Discussion Using flow cytometric analysis of the hypodiploid peak, both CP20 and DFO induced apoptosis of HL60 cells causing 52.7+3.8% and 60.2+4.5% apoptotic death 127 respectively at 24h, similarly, hydroxyurea caused a significant induction of apoptosis compared to control, 57.5±7.5% and 6.5±0.7% apoptosis respectively, P<0.001 (Figure 4.6). Using flow cytometric analysis of the cell cycle (Figure 4.1), both the chelators, CP20 (G(/G|-93.6%, S-2.7% and G2/M-3.6%) and hydroxyurea (Go/G]-89.9%, S-7.3% and G2/M-2 .7 %) caused cell cycle arrest in late Gj by 24h, compared with control (Gq/G,- 69.2%, S-15.2% and G/M-15.6%), p<0.001 (Table 4.1a). These results show that after a 24h incubation with either the iron chelators or hydroxyurea there is a significant increase in the amount of apoptosis compared to control, therefore in the next section, investigation as to the rate of apoptotic induction in the HL60 cell line following addition of chelators or hydroxyurea was examined. Furthermore, the effect of iron chelators on HL60 cells as shown in Figure 4.7 and Table 4.1a are in agreement with other authors (Hileti et a i, 1995; Fukuchi et al., 1994). Like the chelators, HU was also found to induce a significant amount of apoptosis and cell cycle arrest compared to control and this may be explained by the effect of HU on ribonucleotide reductase. Whether the anti-proliferative effect of the chelators in this cell type is through direct inhibition of ribonucleotide reductase however needs further experimentation. 43.2.2 Effect of time and concentration on apoptosis and the cell cycle o f HL60 cells following chelator or hydroxyurea treatment It was important to examine the conditions under which the iron chelators arrested cell cycle progression since a previous study by Rentonet a l, 1996 suggested that iron chelators arrested proliferation at different stages of the cell cycle depending on the concentration and duration of drug exposure. Hoyes et a l, 1993 showed that using low concentrations of chelators, <11 pM IBE, cells accumulated in the G2/M phase of the cell cycle. Different in vitro actions of chelators depending on their duration of exposure and concentration in a cell may have ramifications for the successful application of iron chelator therapy in vivo. Experimental procedure HL60 cells, 2x10^ cells/ml in RPMI-1640 medium were incubated at 37°C/5% CO2 for either; a. 24h with varying concentrations of CP20 or DFO (300, 100, 33, 1 IpM IBE) or b. for varying times (1, 2, 4, 6 , 8 or 24h) at constant concentrations of CP20 or DFO (300pM IBE) or HU (ImM). Cells were then analysed for cell cycle and apoptosis by flow cytometry, section 2.4.1 128 C P 20 Figure 4.6: Induction of apoptosis in HL60 cells by iron chelators and hydroxyurea HL60 cells were incubated for 24h with either PBS, chelators (300pM IBE) or HU (ImM). The cells were washed, fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 6 independent experiments done in duplicate. 129 80 HU 60 - CP20 DFO <2 o Q. 40 - & < 2 0 - , PBS 0 1 0 20 3 0 Time (Mrs) Figure 4.7a: Effect of incubation time on chelator and hydroxyurea-induced apoptosis in HL60 cells HL60 cells were incubated with 300|liM IBE chelators or ImM HU for either 0, 1, 2, 4, 6 , 8 or 24h. Control cells received an equal volume of PBS. The cells were washed, fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 130 SAMPLE INCUBATION % CELLS IN PHASE TIME (hr) GO/Gl S G2/M Control 1 73.413.6 15.111.4 11.511.8 CP20 1 68.714.5 14.912.2 16.412.1 HU 1 70.012.2 16.711.3 12.613.1 DFO 1 67.715.6 14.712.3 17.511.7 Control 4 70.312.6 15.513.5 14.212.2 CP20 4 76.711.6 12.013.2 1 1 .2 1 0 . 8 HU 4 75.113.9 14.212.4 10.613.1 DFO 4 72.212.7 14.511.8 13.211.2 Control 6 72.113.3 15.012.1 13.011.8 CP20 6 87.912.5** 10.611.6* 1.610.7*** HU 6 89.312.6** 8 . 1 1 2 .2 ** 2.610.9*** DFO 6 87.712.9** 10.510.6* 1.810.5*** Control 24 69.212.4 15.211.6 15.611.5 CP20 24 93.613.9*** 2.710.8*** 3.610.9*** HU 24 89.912.2*** 7.311.0** 2.710.8*** DFO 24 79.914.1** 12.912.3 7.912.1** Table 4.1a: Effect of duration of exposure of iron chelators on the HL60 cell cycle . HL60 cells were incubated with control, PBS CP20 and DFO (300|xM IBE) or HU (ImM) for either 1, 4, 6 or 24h. The percentage of cells in each cell cycle phase was assessed by flow cytometry after staining with propidium iodide. The results are expressed as the mean ± SD of 3 independent experiments in duplicate. Significant differences in the cell cycle kinetics between control, chelator and hydroxyurea treated cells were identified at the p<0.05 (*), p<0.01 (**) and p<0.001 (***) levels using the students T test. 131 Results Time dependence of exposure to chelators on apoptosis and cell cycle arrest. Examination of the effect of incubation time on chelator-induced (300|aM IBE) or hydroxyurea-induced (ImM) apoptosis in HL60 cells showed that a significant increase in apoptosis as early as 6 h of incubation with CP20, DFO or HU (12.2±2.0%, 14.6±0.2% and 11.8 ± 0.8% respectively) compared to control (3.3±0.4%), p<0.01(Figure 4 .7 a ). Cell cycle arrest could also be demonstrated by flow cytometry as early as 6 h after the addition of either the iron chelators (300|iM IBE) or HU (ImM) compared with control, p<0.001 (Table 4.1a). Further experiments of ‘pulse-chase’ design were performed to investigate the evolution of apoptosis with time following a6 h pulse with chelators or hydroxyurea. HL60 cells were incubated for 6 h at 37°C/5% CO2 with either PBS, CP20, DFO (300jiM IBE) or HU (ImM). After this time, the samples were washed x3 in culture medium, re suspended in Iml of RPMI-1640 medium and incubation allowed to continue for a further 24, 48 or 96h prior to analysis by flow cytometry. Analysis at 24h revealed significant apoptosis in chelator-treated (41.4±3.8% and 49.2±2.6% for CP20 and DFO respectively) and HU-treated cells (56.7+4.9%) compared to control (5.1+0.9%), p<0.001. Analysis at 48h showed that 100% of chelator and HU- treated cells were apoptotic compared to control (6.6+2.1%) (Figure 4.7b). Concentration dependence of exposure to chelators on apoptosis and cell cycle arrest. Having established that an effect on apoptosis and cell cycle as early as 6 h, using chelator concentrations of 300pM IBE it was important to determine at what concentration apoptosis and cell cycle arrest occurs. HL60 cells were incubated with either PBS and various concentrations of iron chelators (11|O.M-300| llM IBE) for 24h. A concentration of 33|LiM IBE was shown to be sufficient to induce a significant amount of apoptosis after 24h with CP20 or DFO (22.2+0.5% and 20.3±1.0% respectively) compared to control (3.8±0.3%), p<0.001 (Figure 4.7c). Cell cycle arrest was also shown to occur using an iron chelator concentration of 33|iM IBE and using a chelator concentration of lljxM IBE the cells were still in cycle (Table 4.1b). There were no significant differences observed between the abilities of CP20 and DFO at various concentrations to affect the cell cycle. Discussion Incubation of HL60 cells with the chelators caused apoptosis (Figure 4.7a) and cell cycle arrest in a dose dependent manner. However, there were no differences observed 132 120 100 o(/> OQ. a < CP20 Time Mrs Figure 4.7b: Effect of a short puise of chelators and hydroxyurea on apoptosis in HL60 cells HL60 cells were incubated with 300)iM IBE chelators or ImM HU for 6 h. Control cells received an equal volume of PBS. The cells were washed x3 and resuspended in culture medium. After a further 24, 48 or 96h, aliquots of 1ml were removed and the cells were fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 133 70 60 DFO CP20 50 40 o 30 OQ. Q. < 20 10 PBS 0 0 100 200 3 0 0 4 0 0 Chelator Gone. IBE Figure 4.7c: Effect of chelator concentration on apoptosis in HL60 cells HL60 cells were incubated with either control (PBS) or chelators, CP20 and DFO at concentrations of 0, 11, 33, 100 or 300|iM IBE for 24h. After this time the cells were fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean of 3 independent experiements done in duplicate. 134 SAMPLE Cone. \iM IBE % OF CELLS IN PHASE GO/Gl S G2/M Control 72.714.4 13.312.2 13.915.7 CP20 300 92.611.1*** 4.812.1** 2.610.9*** 1 0 0 93.113.8*** 4.612.3** 2.311.0*** 33 89.612.2*** 5.110.6** 5.311.9*** 11 73.113.8 11.611.4 15.214.8 DFO 300 93.512.3*** 3.912.1** 2.610.4*** 1 0 0 92.614.1*** 4.110.6** 3.311.3*** 33 92.113.9*** 5.113.8* 2.810.2*** 11 77.914.1 1 1 .2 1 2 . 0 10.612.1 Table 4.1b: Effect of chelator concentration on the cell cycle status of HL60 cells HL60 cells were incubated with control or the chelators at concentrations of 11, 33, 100 or 300|iM IBE for 24h. The percentage of cells in each phase of the cell cycle was assessed by flow cytometry after staining with propidium iodide. The results are expressed as the mean ± SD of 4 independent experiments done in duplicate. Significant differences in the cell cycle kinetics between control and chelators were identified at the p< 0.01 (*), p<0.05 (**) and p<0.001 (***) levels using the paired students T test. 135 between the relative abilities of either the iron chelators, CP20 or DFO or HU to induce apoptosis in the HL60 cell line. The HL60 myeloid cell line is clearly sensitive to iron deprivation by CP20 and DFO and inhibition of the tyrosyl radical of RR by HU, inducing apoptosis after a6 h incubation with either HU or the chelators (Figure 4.7b) and at a chelator concentration of 33pM IBE (Figure 4.7c). Hoyes et al., 1993 showed that incubation of K562 cells with low concentrations of chelators, <33|xM IBE caused an accumulation of cells in the G2/M phase of the cell cycle. However, in this section low chelator concentrations (ll|xM IBE) were similar to control and the cells continued to proliferate as normal (Table 4.1b). The human promyelocytic leukaemia line, HL60 established by Collins et al., (1977) consists predominantly of promyelocytes (85%) and a small fraction of more mature myeloid elements that can be efficiently increased by treatment with dimethylsulphoxide, butyric acid or dimethylformamide (Collins et a l, 1978). HL60 cells continuously proliferate in suspension culture with a doubling time of 36-48h, unlike other cell lines which have a doubling time of 12-24h, this characteristic alone makes HL60 cells unusual (Collins et a l, 1987). Cell cycle data revealed that with the addition of either CP20, DFO or HU, a proportion of the cells could also be arrested as early as 6 h incubation (Table 4.1a) with the drugs, suggesting a possible association between the ability of the iron chelators and HU to induce apoptosis and arrest cell proliferation. Several lines of evidence indicate that some molecules which trigger apoptosis are also involved in cell proliferation, indeed many transcription factors (cMyc) and tumour suppressor proteins (pRb and p53) exert direct effects both on cell viability and on cell cycle progression. 4.3.2.3 Analysis of the point in cell cycle at which cells are committed to apoptosis by iron chelators and HU: Bromodeoxyuridine (BrdU) analysis of chelator treated HL60 cells The previous section showed that cell cycle arrest and apoptosis are induced rapidly in HL60 cells. To show at which point in the cell cycle these cells become susceptible to the effects of iron chelation, the 5-bromo-2'deoxy-uridine (BrdU) assay, was used as an alternative to the 3H-thymidine uptake assay. The principle is that BrdU can be incorporated into DNA in place of thymidine and using a primary monoclonal antibody directed against BrdU (Vanderlaan et a l, 1985), cells which have incorporated BrdU can be detected with a second fluorochrome-conjugated antibody to the primary antibody. 136 Experimental procedure The human leukaemic HL60 cell line was used to investigate the effect of iron chelators on DNA synthesis in proliferating cells. HL60 cells were used at a concentrations of 2x10^ cells/ml and incubated with the iron chelators, CP20 and DFO 300pM IBE and hydroxyurea at ImM for 6 h, a time point which in the previous section was shown to be sufficient to cause a significant increase in apoptosis. The cell suspension was centrifuged at 1500rpm for 10 minutes and the supernatant aspirated. BrdU labelling medium (Boeringer Mannheim, Germany) at a final concentration of lOpmol/l BrdU in sterile culture medium was added and the cells resuspended. The cell suspension was incubated for Ih at 37®C, 5% CÜ 2 and after this time cells were washed x2 in 1ml PBS. There was no BrdU labelling medium incorporated in the negative control, which instead contained 1ml of PBS. Cells were then fixed in 1% paraformaldehyde (stored at 4°C, but used at RT.) for 5min, followed by a wash in PBS, then fixed and permeabilized in 70% methanol (stored at -20°C) overnight. Cells were then washed in PBS and resuspended in a solution of 0.2mg pepsin in 1ml 2M HCl, pH 1.5 (pepsin in HCl is used to punch holes in the nucleus to enable the anti-BrdU antibody to get to the incorporated BrdU) at room temperature for Ih (Kirkhus & Clausen, 1990). This solution was then neutralised with 1ml O.IM borax and washed x2 with PBS. To the cell pellet, lOOp.1 anti-BrdU antibody (DAKO, diluted 1:20 in PBS/1 % PCS) was added and the cells left at room temperature for 30mins. This was then washed in PBS/1% PCS before the addition of 100|il anti-mouse Ig fluorescein- conjugated secondary antibody (DAKO, diluted 1:10 PBS/1 % PCS), was added and incubated for 30min at room temperature. Cells were then washed in PBS/PCS and resuspended in a 1ml cell suspension containing 1 mg/ml RNase and 50|ig/ml propidium iodide, and analysed by flow cytometry (section 2.4.1). Demonstration of BrdU incorporation and bivariate cell cycle analysis in HL60 cells . Experimental design. HL60 cells were incubated with PBS, chelators (300|iM) or HU (ImM) for 6 h, washed and then incubated with BrdU for a further Ih. Bivariate analysis was undertaken by flow cytometry for cell cycle (red fluorescence , horizontal axis) and of BrdU positivity (green fluorescence, vertical axis) (Figure 4.8). 137 i 20 30 40 50 60 PMT4 Figure 4.8: BrdU analysis of the cell cycle of HL60 cells HL60 cells were incubated with BrdU for Ih, washed and fixed in 70% methanol. The cells were then incubated for a further 30 mins with a BrdU antibody and analysed by flow cytometry. Boxes C and D represent BrdU -ve and +ve stained cells respectively as measured by increasing green fluoresence (PMT2). 138 Results. Figure 4,8 shows the bivariate DNA-BrdU histogram generated by HL60 cells incubated with BrdU for Ih. Gi and G 2/M populations have only a background green fluorescence (FITC). On the horizontal axis (linear scale) incorporation of propidium iodide is shown as red fluorescence and this distribution gives an analysis of cell cycle profile. The vertical axis shows relative incorporation of BrdU as shown by green fluorescence (FITC, log scale). The proportion of cells BrdU positive cells in each phase of cell cycle can be inspected by eye. It can be seen that cells incubated with PBS only, S-phase cells have a green fluorescence (as measured by an increase in PMT2 on the vertical axis. Figure 4.8) and produce a horseshoe-shaped pattern with mid-S-phase cells having the highest fluorescence Effect of Ih incubation with BrdU followed by incubation with HU or chelators on cell cycle and apoptosis Experimental design. HL60 cells were pulse labelled with BrdU Ih and then incubated continuously for 1,4, 6 , 8 or ISh with either control, chelators (300pM IBE) or HU (ImM). Bivariate analysis was undertaken by flow cytometry for cell cycle (red fluorescence, horizontal axis) and of BrdU positivity (green fluorescence, vertical axis). On the horizontal (linear scale) incorporation of propidium iodide is shown as red fluorescence and this distribution give an analysis of cell cycle profile. The vertical axis shows relative incorporation of BrdU as shown by green fluorescence (FITC, log scale). The proportion of cells BrdU positive cells in each phase of cell cycle can be inspected by eye and were also analysed for by gating in the boxes shown (Figures 4,9). Pattern of BrdU incorporation in control cells In control cells (PBS), S-phase cells that incorporate BrdU show the typical green fluorescence of FITC after Ihour incubation (Figure 4.9a) This same cohort of S-phase cells progress through the cell cycle, with the fluorescence beginning to appear in the G2/M cells by 8 h (Table 4.2). By 18h, the cohort of fluorescence had progressed through the cell cycle and appear in Gq/Gi(Figure 4.9a and Table 4.2). Table 4.2 compares the percentage of cells either staining with BrdU (BrdU -fve) or not (BrdU -ve) in each phase of the cell cycle (apoptosis, Gq/G], S and G2/M phase). 139 Pattern o f BrdU incorporation in CP20 treated cells. When cells are incubated with CP20, incorporation of BrdU into S phase cells was not significantly affected at Ih 18.8+3.6% compared with control of 21.2+1.1%. At 4h 15.5±4.1% of S phase cells are BrdU positive compared with 20.2±3.6% in control. By 6, 8 and 18h, the percentage of BrdU positive cells in S-phase had significantly decreased (6.8±2.1%, 4.8±3.3% and 1.9% respectively) compared to control at 6, 8 and 18h (15.0±3.1%, 10.1±4.8% and 8.5±2.1% respectively), p<0.05 from two experiments done in duplicate. The histograms (Figure 4.9b) also show a population of cells to the left of the brightly stained BrdU S-phase peak by 6h indicating that cell treated with CP20 had apoptosed from the S-phase of the cell cycle.Table 4.2 shows that there was a significant decrease in S-phase cells compared to control as indicated above, p<0.05. After 6h, the percentage of cells apoptosing was 21.0% (BrdU +ve) and 10.1% (BrdU -ve) compared to control at 6h 3.4% (BrdU +ve) and 10.7% (BrdU -ve), once again indicating that the majority of cells apoptosing after incubation with CP20 come from the S-phase of the cell cycle. Pattern o f BrdU incorporation in HU treated cells. Cells were also incubated with hydroxyurea for 1,4, 6, 8 or 18h, prior to analysis of BrdU incorporation. The percentage of BrdU positive cells in S phase at 1 and 4h were 17.6+2.6% and 15.2+2.2% respectively. As with the results for CP20, HU treated cells were found to cause a significant decrease in the proportion with incorporation of BrdU by 6h (7.5±3.4% of cells in S-phase) compared to control, (15.0+3.1%) p<0.05. As with the iron chelators, apoptosing cells were likely to come from the S-phase and Gq/Gi phase of the cell cycle (Figure 4.9c). In Table 4.2, apoptosis as outlined previously was evident after 6h incubation, 17.8% (BrdU +ve) and 11.8% (BrdU -ve), after a 18h pulse chase with HU, apoptosis had increased to 27.7% (BrdU -t-ve) and 19.3% (BrdU -ve). Pattern o f BrdU incorporation in DFO treated cells. When cells were incubated with DFO (300)J,M IBE), a change in the pattern of incorporation of BrdU was again evident. The percentage of cells in S-phase at Ih and 4h were 19.3+5.2% and 16.5±2.4% respectively. However the percentage of cells in S-phase had only significantly decreased after a 18h pulse (2.1±2.7% compared to control 8.5± 2.1%), p<0.01. This contrasts with CP20 where changes were seen by 4-6h. Again the histogram suggests that apoptosing cells are recruited from the S-Phase of the cell cycle (Figure 4.9d). Table 4.2 shows that after a 6h pulse, the percentage of cells apoptosing was 3.4% (BrdU 4-ve) and 12.7% (BrdU -ve) indicating that in comparison to the effects 140 Oh Ih § i t I I I 10 20 30 40 50 60 10 20 30 40 50 60 PMT4 PMT4 4h 6h g § 10 20 40 5 0 60 0 10 20 30 40 50 60 PMT4 8h 18h § 20 30 40 50 60 PMT4 PMT4 Figure 4.9a; Cell cycle kinetics of HL60 cells: BrdU analysis HL60 cells were incubated with BrdU for Ih, followed by incubation with PBS for Oh ,lh, 4h , 6h , 8h and 18h . After this time the cells were washed and incubated for 3()mins with a BrdU antibody and analysed by flow cytometry. 141 Oh Ih 3 3 20 30 40 50 60 10 20 30 40 50 68 PMT4 PMT4 4h 6h 3 3 10 20 30 40 50 60 0 10 20 30 40 50 60 PMT4 PMT4 8h 18h 3 3 20 30 40 50 60 10 20 30 40 50 60 PMT4 PMT4 Figure 4.9b: Effect of CP20 on the cell cycle kinetics of HL60 cells: BrdU analysis HL60 cells were incubated for Ih with BrdU, followed by a Oh ,lh , 4h , 6h , 8h and 18h incubation with CP20 (300|j,M IBE). After this time cells were incubated for 30mins with a BrdU antibody and analysed by flow cytometry. 142 Oh Ih § I 20 30 20 30 40 50 60 PMT4 PMT4 4h 6h § § 10 20 30 40 50 60 20 30 40 50 60 PMT4 PMT4 8h 18h § § I 20 30 40 50 60 20 30 40 50 60 PMT4 PMT4 Figure 4.9c: Effect of Hydroxyurea on the cell cycle kinetics of HL60 cells BrdU analysis. HL60 cells were incubated with BrdU for Ih followed by a Oh ,lh , 4h , 6h, 8h and 18h incubation with HU (ImM). After this time cells were washed and incubated with a BrdU antidody and analysed by flow cytometry. 143 Oh Ih § I 10 20 30 40 50 6 0 20 30 40 50 60 PMT4 PMT4 4h 6h CN-! 10 20 30 40 50 60 10 20 30 40 50 60 PMT4 PMT4 8h 18h § 3 P! E 0 10 20 30 40 50 60 20 30 40 50 60 PMT4 PMT4 Figure 4.9d: Effect of DFO on the cell cycle kinetics of HL60 cells: BrdU analysis HL60 cells were incubated with BrdU for Ih followed by a Oh ,lh , 4h , 6h , 8h and 18h incubation with DFO (300piM IBE). After this time cells were washed and incubated with a BrdU antibody for 30mins and analysed by flow cytometry. 144 Sample BrdU status % of cells in phase of cell cycle Apop G O /G l S G 2/M PBS Oh + 0.0 4.2 12.9 3.3 - 3.3 74.5 - 1.6 PBS Ih + 0.1 1.3 21.2 3.3 - 5.6 60.8 - 1.6 CP20 Ih + 0.0 2.0 18.8 5.6 - 6.1 57.8 - 5.4 HU Ih + 0.1 3.8 17.6 7.3 - 6.0 62.4 - 7.0 DFO Ih + 0.0 1.3 19.3 5.8 - 4.9 60.0 - 7.9 PBS 6h + 3.4 2.5 15.0 6.8 - 10.7 58.9 - 2.7 CP20 6h + 21.0 5.0 6.8 2.5 - 10.1 51.7 - 2.9 HU 6h + 17.8 6.4 7.5 3.2 - 11.8 51.7 - 1.6 D F0 6h + 3.4 3.8 14.6 7.2 - 12.7 56.2 - 2.1 PBS 18h + 5.2 23.3 8.5 5.1 - 11.2 41.1 - 5.7 CP20 18h + 36.8 3.3 1.9 1.7 - 16.5 40.6 - 1.1 HU 18h + 27.7 3.1 1.6 1.5 - 19.3 46.1 - 0.7 DFO 18h + 18.7 7.1 2.1 2.8 - 19.4 48.9 - 1.0 Table 4.2: Effect of pulse-chase of chelators or hydroxyurea on the cell cycle of HL60 cells: BrdU analysis. HL60 cells were incubated with BrdU for Ih, followed by a pulse chase for either 0, 1, 6 or 18h with PBS, chelators (300|liM IBE) or HU (ImM). After this time the cells were washed and incubated for a further 30mins with a BrdU antibody. The effect of the drugs on the cell cycle, comparing +ve and -ve BrdU stained cells was analysed by flow cytometry. 145 with CP20, the majority of cells apoptosing after incubation with DFO come from the Gq/G i population of cells. After a 18h pulse chase with DFO the percentage of cells apoptosing were 18.7% (BrdU +ve) and 19.4% (BrdU -ve). Discussion. The findings in this section elucidate several points about how chelators and HU affect cell cycle, DNA synthesis and induce apoptosis. Firstly the results show that new DNA synthesis is inhibited by both chelators and HU resulting in reduction in the proportion of BrdU positive cells in S phase. Secondly the results show that this inhibition is significantly faster with CP20 and HU than with DFO. These findings are consistent with known slower uptake of DFO into cells and subcellular compartments (Hoyes et al., 1993) as well as the known slower inhibition of enzymes such as ribonucleotide reductase (C oopérera/., 1996). The finding that apoptotic cells are mainly recruited from the BrdU positive fraction is supportive of the concept that cells synthesising DNA at the time of exposure to chelators are susceptible to apoptotic induction. The finding that the pattern of BrdU incorporation into cycling and apoptotic cells is broadly the same for CP20 and HU is consistent with the idea that inhibition of RR is a key mechanism for apoptotic induction by CP20. 4.4 Comparison of apoptotic induction in haemopoietic cells by hydroxyurea and iron chelators. 4.4.1 Rationale Micromolar concentrations of iron chelators have been previously shown to induce a dose-dependent inhibition of proliferation in murine haemopoietic progenitorsin vitro (Porter et a l, 1989; Hoffbrand et at., 1991; Hoyes et at., 1992). However the exact cell type or stages in cell differentiation when this occurs have not been elucidated. The mechanism by which this anti-proliferative phenomenon occurs is also unclear although it has been assumed by many investigators to be secondary to inhibition of RR. This is important because CP20, in addition to causing a dose dependent marrow hypoplasia in mice (Hoyes er fl/., 1993; Porter gr a/., 1991), causes agranulocytosis and neutropenia in a proportion of humans taking the drug (Wonke et a l, 1998). Results in chapter 3 indicated that the effects of the iron chelators on haemopoietic progenitors were cell cycle dependent. The purpose of this section therefore has been to examine the stage(s) in the differentiation of haemopoietic cells at which they become susceptible to apoptosis by iron chelators and to compare this with the effect of 146 hydroxyurea, which is thought to have actions deriving from relatively specific inhibition of RR. This has been achieved by comparing the effects of chelators and hydroxyurea on apoptosis of human CD34^ cells as they differentiate during continuous culturein vitro with SCF, EL3 and IL6 as well as by comparing such effects with those on a relatively committed population of murine granulocytic cells. 4.4.2 Effect of hydroxyurea and iron chelators on apoptosis in committed murine granulocytes. R ation ale. A monoclonal rat neutrophil antibody was used to identify relatively mature granulocytes (promyelocytes to mature granulocytes) (Hirsch et a i, 1983) freshly isolated from the bone marrow of mice, and to examine the extent of apoptosis after the addition of either the iron chelators or HU in the fraction staining positive with this antibody. The rat monoclonal antibody (Ab) 7/4 produced against neutrophil rich cultured bone marrow populations defines a polymorphic neutrophil differentiation antigen (Ag). Ag 7/4 expression was characterised on cells from C57BL6 mice using FACS analysis. Experimental procedures Male C57bL/J mice aged 10-12 weeks were sacrificed by cervical dislocation and progenitor cells were flushed from the femur cavity with 5ml of RPMI-1640 medium supplemented with 10% FCS and 2% Pen/Strep, using a 25G hypodermic needle (as previously described, Hoyes et a l, 1993). The isolated cells were washed x2 with 5ml of RPMI-medium and resuspended at a concentration of 2x10^ cells/ml. After 24h incubation at 37°C with either PBS, CP20, DFO (300pM IBE) or HU (ImM),the cells were pelleted and fixed in 70% ethanol for 60mins. After this time, cells were washed x3 with 3% BSA/PBS. A rat anti-mouse primary neutrophil antibody (Serotec, Oxford, UK), recognising promyelocytes to mature granulocytes was added at a dilution of 1:4 or a rat IgG2a (Serotec, Oxford, UK). A rat IgG2a negative control , . (Serotec, Oxford. UK) was incorporated as a negative control at a final concentration of lOjig/ml. Cells were then placed on ice for 30mins. After this time, cells were once again washed 3x with 3% BSA/PBS. The secondary antibody, a F(ab')2 rabbit anti-rat IgG: FITC (Serotec, Oxford, UK) was added at a dilution of 1:50 from original stock and again left on ice for 30mins. After three washes in 1ml PBS, cells were resuspended with 0.25ml of 3% BSA/PBS, 0.25ml RNase, 1 mg/ml and 0.5ml propidium iodide, 50pg/ml. Cells were left in the dark for 24h at 4®C prior to the analysis of apoptosis by flow cytometry 147 (section 2.4.1). During flow cytometric investigation, cells staining positive for the antibody (FITC incorporation as measured on the PMT2 parameter of the FACS machine) were gated and further analysed for apoptosis by PI incorporation (as measured on the PMT3 parameter of the FACS machine) as shown in Figure 4.10. Results and discussion Figure 4.11 compares the effect of iron chelators and hydroxyurea on apoptosis in murine granulocytic cell types (promyelocytes to mature cells) as recognised by specific antibodies for promyelocytes differentiating to granulocytes. In promyelocytes to mature cells, the finding that hydroxyurea induces a 2-fold increase in apoptosis over control (p<0.01) in the population of cells staining positively with the antibody suggests that ribonucleotide reductase may be a target enzyme for apoptotic induction in a proportion of this heterogeneous population of cells. Despite the small proportion of proliferating cells, there was an obvious 2-fold increase in apoptosis with HU. Surprisingly, in contrast the iron chelators, CP20 and DFO had no increased induction in the amount of apoptosis in the granulocytic lineage compared to control. It is not clear why HU should induce apoptosis in this relatively mature population of granulocytes but not CP20 or DFO. After the promyelocyte stage, myelocytes and neutrophils do not divide and therefore it would be surprising if either the chelators or HU induced apoptosis. The apoptosis seen with HU is therefore likely to be occurring in promeylocytes. The lack of effect with chelators could simply be a dose dependent effect, with higher concentrations being required. However similar concentrations of HU and chelators in other cell types (see section 4.3 & 4.4.3) showed broadly similar degrees of apoptosis with either HU or chelators. One possible explanation for the differential sensitivity to HU and chelators is that promeyelocytes being the last division in granulocyte differentiation have made all the RR necessary for this final division. HU by virtue of its action in inhibiting RR which has already been made by scavenging the tyrosyl radical could still inhibit DNA synthesis and potentially induce apoptosis. By contrast the action of the chelators may primarily be to inhibit RR, either when iron is being incorporated into the molecule during synthesis, or is necessary during regeneration of the tyrosyl radical after it is consumed during ribonucleotide reduction (Fontecave, 1998). Therefore in cells undergoing terminal divisions, the inhibition of new RR synthesis or regeneration may be insufficient to induce apoptosis whereas the inactivation of enzymes which is already fully functional by HU could in principle be sufficient to promote apoptosis. 148 Increasing green fluorescence C O u N T PMT2 LOG B. Increasing red fluorescence GO/Gl C O u N T Apoptosis G2/M PMT3 LOG Figure 4.10: Measurement of apoptosis in murine neutrophils Murine neutrophils and their immediate precursors were selected using a monoclonal rat anti-neutrophil antibody (FITC, Green fluorescence, PMT2 LOG). Cells staining positive with the antibody (A) were gated and analysed for % apoptosis after PI staining (red fluorescence, PMT3 LOG)(B). 149 . •<. CP20 DFO Figure 4.11: Effect of iron chelators and hydroxyuea on murine neutrophils and their immediate precursors Murine bone marrow was extracted from the femurs of 10-12 week old Balb-C mice. 1x10^ cells/ml were incubated with either PBS, the chelators (300pM IBE) or HU (ImM) for 24h. After this time, neutrophils and their immediate precursors were selected using a monoclonal rat anti-neutrophil antibody followed by PI staining for apoptosis measurement by flow cytometiy. The data shown are the mean ± SD of 8 independent experiments done in duplicate. 150 4.4.3 Effect of hydroxyurea and iron chelators on apoptosis and cell cycle in CD34+ cells and their progeny in continuous culture. 4.4.3.1 Relative effect of continuous exposure of to chelators or hydroxyurea on apoptosis of CD34+ cells. R ationale In Chapter 3, it was established that CP20 and DFO caused an increase in apoptosis in CD34^ cells compared to control after 3 days in culture with the growth factors, SCF, IL3 and IL6. Previous work using the same system of culture in the same laboratory showed that after 12h, CD34^ cells are in the Go phase of the cell cycle and enter the cell cycle, approximately 48h after the addition of IL3, IL6 and SCF (Williams et aL, 1997; Tiwari, 1998). The purpose of this section was therefore to investigate whether exposure to HU caused a similar degree of apoptosis in CD34^ cells and their progeny as chelators (Section 3.4). In the following section, the percentage of cells induced to undergo apoptosis by the iron chelators (300|llM IBE) and HU (ImM) are compared. Furthermore, in section 4.4.2 the results suggested that murine neutrophils and their immediate precursors were susceptible to HU but not iron chelators. In this section the stage of differentiation in human haemopoietic cells which is susceptible to cell cycle arrest by HU and chelators is examined. Experimental procedure CD34^ human haemopoietic progenitor cells were cultured and isolated as indicated in section 2.3.7. 2x10^ cells/ml were set up in culture with relevant growth factors (IL3, IL6 and SCF) (section 2.3.7) and incubated continuously with either PBS, the iron chelators, CP20 or DFO (300pM IBE) or HU (ImM) for 7 days at 37®C/5%C02 unless otherwise stated. Aliquots of 1ml were removed on days I, 3, 5 or 7 and analysed for apoptosis and cell cycle by flow cytometry as outlined in section 2.4.1 R esu lts In Figure 4.12a. it can be seen that by day 3 of continuous culture of CD34^ cells with CP20 or hydroxyurea, both drugs had induced apoptosis in the cells compared to control. Hydroxyurea induced significantly more apoptosis, 66.1±1.8%, compared to control 6.9+1.3%, p<0.001. The same proportions of increased apoptosis are seen at day 5 and day 7. Furthermore HU caused more apoptosis than the iron chelators, p<0.01. Results from chapter 3 indicate that by day 3, the CD34^ cells were proliferating and 151 SCF+IL3+IL« 100 8 0 PBS (0 6 0 w o CP20 4-^ o. e Q. < HU 4 0 DFO 20 • DAY 1 DAY 3 DAY 5 DAY 7 F ig u re 4.12a: Effect of continuous exposure of CD34^ cells to iron chelators and hydroxyurea on apoptosis PBS, chelators (300pM IBE) or HU (ImM) were added 2h after isolation of CD34+ cells (1x10^ cells/ml), with aliquots of 1ml taken on the days in question. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement. The data shown are the mean ± SD of 3 independent experiments done in duplicate. 152 were committed to the myeloid lineage after incubation with the growth factors IL3, IL6 and SCF using morphological and cell antigen expression analysis. Neither HU or the chelators increased apoptosis on day 1 at a time when the cells were quiescent. 4.4.3.2 Effect of continuous exposure of CD34+ cells to chelators and hydroxyurea on cell cycle Rationale and experimental procedure. In order to interpret the actions of HU and iron chelators on apoptosis, the cell cycle of treated and untreated cells was examined at days 1,3,5 and 7. R e su lts. Hydroxyurea, as well as inducing apoptosis as shown in the previous section was also shown to cause cell cycle arrest. Table 4.3. Similarly both the iron chelators, CP20 (S-phase-4.0±l.l%, G2/M-phase-4.4±0.8%) and DFO (S-phase-4.2±0.9%, 2 G/M-phase- 2.9±0.5%) caused cell cycle arrest by day 3 compared to control PBS (S-phase- 19.8±1.5%, G2/M-phase-16.5±1.3%). Discussion. As outlined in chapter 3 and shown in Figure 4.12a, the chelators also caused significant apoptosis, although this is a little less than that seen with HU at the concentrations used. Indeed HU induces more apoptosis than DFO and CP20 on days 3, 5,and 7 of continuous exposure. There are however no consistent differences between HU and the chelators on cell cycle but since by day 5 the majority of cells have apoptosed (section 3.3), analysis of cell cycle in the residual cells may not reflect a representative picture of the relationship between cell cycle and apoptosis. In order to reveal the relationship between cell cycle and apoptosis more clearly, a shorter exposure to HU or chelators after different periods of CD34^ culture would be required. This is examined in the section below. 4.4.4 Effect of 24h exposure of CD34+ cells to hydroxyurea and chelators on apoptosis and cell cycle. In the above experiments it was established that after continual exposure of CD34^ cells to both hydroxyurea and iron chelators, apoptosis and cell cycle arrest resulted after 3 days. However this makes it difficult to be certain at what point during cellular 153 SAMPLE DAY % C ELLS IN PH A SE G O /G l S G 2/M Control 1 92.1±1.2 4.1+1.9 3.8+0.7 CP20 1 94.8+1.3 2.6+0.7 2.6+0.4 HU 1 93.2+1.2 3.1+0.9 3.7+1.1 DFO 1 95.1+0.8 2.1+0.8 2.8+0.5 Control 3 63.7±1.7 19.8+1.5 16.5+1.3 CP20 3 91.6+4.3** 4.0+1.1*** 4.4+0.8*** HU 3 87.7+2.6** 7.2+1.6*** 5.0+0.9*** DFO 3 92.9+2.5 4.2+0.9*** 2.9+0.5*** Control 5 66.9+0.3 16.9+1.4 16.1+1.7 CP20 5 89.1+6.1** 7.5+1.2** 3.4+0.9*** HU 5 81.9+2.8*** 9.2+1.7* 8.8+0.8** DFO 5 91.5+1.7** 3.9+0.9*** 4.5+1.0*** Control 7 70.8+1.5 14.6+1.2 14.5+2.4 CP20 7 84.8+4.2*** 8.9+1.1* 6.2+1.2** HU 7 90.6+2.6 4.3+1.1*** 5.1+3.2** DFO 7 91.6+0.5*** 3.8+0.8*** 4.4+1.0*** Table 4.3 Effect of continuous exposure to iron chelators on the cell cycle of CD34+ cells. PBS, Chelators (BOOpM IBE) or HU (ImM) were added on day 1 with samples taken on the days in question. Cell cycle status was assessed by flow cytometry after staining with propidium iodide. Results are expressed as the percentage of cells in each phase of the cell cycle and represent the mean ± SD of 3 independent experiments in duplicate. Significant differences in the cell cycle kinetics between control and chelator and hydroxyurea treated cells were identified at the p<0.01 (**) and p<0.001 (***) levels using the Students T test. 154 differentiation apoptosis was triggered, therefore it was necessary to investigate whether by adding the chelators and HU for 24h, apoptosis and cell cycle arrest could be induced. Experimental Procedure CD34"^ human haemopoietic progenitor cells were cultured and isolated as indicated in section 2.3.7, 2x10^ cells/ml were set up in culture for 1, 3, 5 or 7 days with relevant growth factors, SCF, IL3 and IL6. After initiation of culture on the respective days, the cells were incubated for a further 24h with PBS, CP20 or DFO (300pM IBE) or HU (ImM) prior to flow cytometric analysis. R esu lts In Figure 4.12b, HU can be seen to induce significant apoptosis after 24h incubation on day 3, 39.1+6.1% compared to control 8.9±1.8%. In chapter 3 maximal apoptosis was evident when either CP20 or DFO were added on days 3 and 5 and of interest was that by day 7 of CD34^ proliferation and differentiation the amount of apoptosis had decreased. Similarly in this section hydroxyurea induces maximal apoptosis on days 3 and 5 compared to day 7, this however could be due to the proliferative status of the cell as outlined in chapter 3. Once again shown in the previous section, 24h exposure to HU caused more apoptosis than CP20 on days 3, 5 and 7 of CD34^ differentiation. As shown in Table 4.4, addition of either HU or chelators for 24h also causes significant cell cycle arrest by day 3, (CP20 had an S-phase-12.4±0.6%, G 2/M-phase- 5.5±0.6%, DFO: S-phase-12.5±0.2%, G2/M-phase-6.7±0.6% and HU: S-phase- 9.2+1.7%, G2/M-phase-1.8±0.6%) compared to control (S-phase-25.2+4.8%, G2/M- phase-9.9+1.7%). By contrast in section 4.4.3.2, no differences were found between the iron chelators and HU with respect to their effects on the cell cycle. When CD34'*’ cells were exposed to either chelators or HU for 24h on different days of differentiation, there were significant differences between chelators and HU during S phase, p<0.01. The results therefore show that neither HU or the chelators affect the cell cycle of CD34^ cells on day 1 when the cells were quiescent in Go/G|. However by days 3, 5 and 7, following the addition of HU and chelators, cell cycle is arrested in G, with a concomitant decrease in the S and G2/M phases of the cell cycle. There is no difference between the ability of CP20 and DFO to arrest the cell cycle under these conditions. However HU at concentrations of ImM is more effective than the chelators at 300pM IBE at arresting the cell cycle (Table 4.4). These findings again show that at the concentrations used, HU is more inhibitory of cell cycle and induces mire apoptosis than the iron chelators. The finding that both cell cycle arrest and apoptosis are increased 155 SCF+IL3+IL6 (M 09 ■ PBS (0 ■55 B CP20 □ HU O S DFO Q. < DAY 1 DAY 3 DAY 5 DAY 7 DAY 9 F ig u re 4.1 2 b : Effect of transient exposure of CD34^ cells to iron chelators and hydroxyurea on apoptosis CD34^ cells were set up in culture with SCF, IL3 and IL6, on days 1, 3, 5, 7 or 9 of CD34^ differentiation, aliquots containing 1x10^ cells/ml were removed and either control, chelators (300pM IBE) or HU (ImM) added for 24h. The cells were washed, fixed in cold 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the maen ± SD of 3 independent experiments done in duplicate. 156 SAMPLE DAY % of CELLS IN PHASE GO/Gl S G2/M Control 1 92.1+1.2 4.2+1.3 3.710.9 CP20 1 90.8±1.3 4.9+0.8 4.311.1 HU 1 92.2±1.2 5.1 + 1.2 2.710.8 DFO 1 90.1±0.8 3.9±1.7 6.011.2 Control 3 64.8+3.7 25.2+4.8 9.911.7 CP20 3 82.1±2.1** 12.4±0.6*** 5.510.6* HU 3 88.9±6.1*** 9.2+1.7*** 1.810.6** DFO 3 80.7+2.1*** 12.510.2*** 6.710.6 Control 5 66.0±0.9 19.010.6 14.911.2 CP20 5 88.5+2.3*** 7.510.9*** 3.910.8*** HU 5 92.8±5.2*** 3.811.1*** 3.310.9*** DFO 5 87.8+2.9**** 9.410.9*** 2.311.0*** Control 7 67.5+1.1 19.310.3 13.110.5 CP20 7 88.0+4.5*** 9.512.1** 2.411.0*** HU 7 89.3±9.9*** 7.210.6*** 3.410.7*** DFO 7 84.5+3.0*** 12.310.8** 3.111.3*** Table 4.4 Effect of 24h exposure of iron chelators on the cell cycle of CD34+ cells after initiation of culture on days 1, 3, 5 or 7 Cell cycle was assessed by flow cytometry after staining with propidium iodide following addition of the drugs on the days in question for 24h. Results are expressed as a percentage of cells in each phase of the cell cycle and represent the mean ± SD of 3 independent experiments in duplicate. Significant differences in the cell cycle kinetics between control and chelator and hydroxyurea treated cells were identified at the p<0.01 (*), p<0.05 (**) and p<0.001 (***) levels using students T test 157 relatively with HU compared with the chelators is consistent with there being a more pronounced effect on RR inhibition and DNA synthesis with HU at these concentrations and itîs ndlnecessary, from these experimental data alone, to invoke additional or alternative mechanisms for apoptotic induction. 4.4.5 Effect of chelator concentration on the cell cycle status of CD34+ cells Rationale In order to examine whether the difference between chelators and HU was simply a consequence of the concentrations used, a dose response for the effect of chelators on cell cycle and apoptosis was performed. Results inchapter 3 suggested that there was no difference in the propensity to apoptosis between DFO and CP20 even at low concentrations (llpM ). However there may be differences in the percentage of cells in the various phases of the cell cycle. The purpose of this section therefore, was to compare ceU cycle arrest between different concentrations of CP20 and DFO. Experimental procedure CD34^ human haemopoietic progenitor cells were cultured and isolated as indicated in section 2.3.7. 2x10^ cells/ml were set up in culture for 3 days with relevant growth factors (SCF, IL3 and IL6). After culture on the respective days, the cells were incubated for a further 24h with PBS, CP20 or DFO at increasing concentrations of lljxM -300pM IBE. R esu lts As shown in Table 4.5, both CP20 and DFO caused a significant cell cycle arrest by day 3, using a concentration of either 300 or lOOjxM IBE, as compared to control, p<0.001. At a concentration of 33pM IBE significant cell cycle arrest (Go/G|-90.2+10.2%, S-phase-7.3±2.1%, G2/M-2.5±0.9%) occurred with CP20 compared to control (Gq/Gi- 75.1±6.8%, S-phase-17.5±3.6%, G2/M-7.5±1.6%), p<0.001. However at the same concentration, DFO did not affect the cell cycle compared to control (Gq/Gi-79.6±1 1.2%, S-phase-12.3±1.8%, G/M-8.I±1.2%). Because apoptosis does not increase further above chelator concentrations of lOOpM IBE (section 3.3.6), and cell cycle inhibition also plateaus at lOOpM (Table 4.5), it follows that the maximum effect of chelators was observed at the experiments where 300pM IBE were used, p<0.001. Therefore the smaller effect of chelators compared 158 Sample Cone. pM IBE % of Cells in Phase GO/Gl S G2/M Control - 75.1±6.8 17.5+3.6 7.5+1.6 CP20 300 93.0±9.9*** 6.2±1.0** 0.8±0.1** DFO 300 93.1+5.7*** 5.9±1.6** 1.0±0.2** CP20 100 92.5±3.8*** 6.7±1.8** 0.8±0.1** DFO 100 89.5±11.1** 7.8±1.9** 2.6+0.5* CP20 33 90.2+10.2** 7.3+2.1** 2.5+0.9* DFO 33 79.6±11.2 12.3±1.8 8.1±1.2 CP20 11 78.9±5.5 13.9±3.9 7.1±1.1 DFO 11 79.6+4.5 13.8±2.7 6.5±0.9 T able 4.5: Effect of chelator concentration on the cell cycle status of CD34^ cells Concentrations of chelators (0, 11, 33, 100, 300pM IBB) were added to cycling CD34^ cells (day 3) for 24h. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ±SD of 3 independent experiments done in duplicate. Significant differences in the cell cycle kinetics between control and chelator-treated cells were identified at a p<0.01(*), p<0.05(**) and p<0.001(***) levels using Students T test. 159 with HU, was in previous experiments unlikely to have been not due to suboptimal concentrations of chelators. It is likely therefore that HU is a more efficient inhibitor of RR than the chelators. This could be due to the fact that HU scavenges the tyrosyl radical directly, whereas the chelators act by inhibiting iron incorporation into the molecule (Cooper et aL, 1996). 4.5 General Discussion The purpose of this chapter was to investigate whether the induction of apoptosis by iron chelators was through the inhibition of cellular proliferation. These two are intimately linked as multicellular organisms can be viewed as the products of cell proliferation, differentiation and cell death. Several lines of evidence suggest that apoptosis and proliferation coincide to some degree (Evanet aL, 1995). Firstly cellular proliferation and apoptosis share some morphological features (e.g. loss of cell volume, condensation of chromatin and disassembly of the nuclear lamina). Secondly deregulation of the cell cycle components has been shown to induce cell cycle arrest and may be involved in triggering apoptosis. Additionally some apoptotic cells express abnormal levels of the cell cycle proteins and often contain active Cdc2, the primary kinase active during cellular proliferation. It is known that iron is essential for cell growth and viability and that iron deprivation results in an inhibition in the synthesis of deoxyribonucleotides (Haq et aL, 1995) and as such there has been much interest in the use of iron chelators as anti neoplastic agents. However, steps leading to eventual cell death during iron deprivation by iron chelators are not fully understood. Ribonucleotide Reductase, which requires non-haem iron (III) for generation of the tyrosyl radical necessary for enzyme activity of the M2 subunit, is a potential target enzyme for inhibition by iron chelators. Indeed a recent paper has shown that the tyrosyl radical of RR is rapidly inhibited within 5 minutes incubation with hydroxypiridinones in K562 cells (Cooper et a l, 1996). It was suggested that this may be the mechanism for the anti proliferative effects of iron chelators, as well as their effects on inducing apoptosis. The findings in this chapter suggest that inhibition of DNA synthesis and ribonucleotide reductase (RR) are unlikely to be the primary mechanisms triggering apoptosis in thymocytes as outlined below, but may be involved in the induction of the apoptotic programme in proliferating cell types such as primary haemopoietic CD34^ progenitor cells and human leukaemic cell lines. 160 The findings in this chapter therefore suggest that additional mechanisms are likely to be required for the induction of thymocyte apoptosis by iron chelators. This is because the known RR inhibitor, hydroxyurea, which interacts directly with the tyrosyl radical of RR (Cooper et a l, 1996) had no significant effect on thymocyte apoptosis at a concentration of ImM and only a small increase in apoptosis over a wide concentration range up to 8mM (Figure 4,4a & b) when compared to control. These results are consistent with the small proportion of cells in cycle in thymocytes. Furthermore although CP20 did significantly affect DNA synthesis as measured by thymidine uptake, as indicated in section 4.2.3, the proportion of cells apoptosing (approximately 60%) with CP20 after 24h incubation, far exceeds the proportion of cells in cycle (approximately 5%) at that time. This suggests the involvement of mechanisms other than RR inhibition to account for apoptotic induction by chelators in thymocytes. These will be examined in the subsequent results chapters. Work in this chapter has shown that both the iron chelators CP20 and DFO and HU increase apoptosis compared to control in all the proliferative cell types (HL60's and CD34^ cells) investigated. Furthermore in this study both the chelators and hydroxyurea caused cell cycle arrest and inhibited DNA synthesis (as measured by BrdU incorporation) compared to control. Such findings are also supported by others; Pattanapanyasatet a l, (1992) suggested that CP20 and DFO exerted their anti-proliferative effects on peripheral blood mononuclear cells through chelation of the ferric ion with subsequent inhibition of DNA synthesis through the inhibition of ribonucleotide reductase. In contrast, several authors have also suggested that the chelators inhibit DNA synthesis by inhibiting cyclin- dependent kinase activity, which ultimately affect the pathways that control cell division (Kulp et a l, 1996). Indeed an earlier study by Smith et a l, 1995 showed that a major effect of iron chelators in blocking cell cycle progression may be mediated through inhibition of cyclin A protein and therefore p33cdk2 activity. Therefore inhibition in the transcription of genes which encode proteins which are known to be involved in cell cycle regulation (e.g. c-Myc as previously discussed) cannot be ruled out as contributing factors to the anti-proliferative effects of iron chelators. Cell culture data have demonstrated that transferrin, the major iron transport protein, is a necessary requirement for cellular proliferation and evidence suggests that transferrin supports proliferation by providing iron for critical cellular processes including DNA synthesis. Indeed many investigators have shown that the anti-proliferative effects of the iron chelators can be reversed with the addition of iron (Lederman et a l, 1984; Alcain et a l, 1994. 161 As an alternative to thymidine, cell proliferation was studied by monitoring the incorporation of BrdU. This proved to be useful in two ways. Firstly it provided information on the proliferative character of the HL60 cell population with a control group of cells passing through the cell cycle by 18h (Figure 4.9a) and secondly it provided knowledge of the fraction of cells at risk of apoptosis from iron chelators. Both the iron chelators, CP20 and DFO and HU were shown to cause an arrest of HL60 cells in the S phase of the cell cycle after 6h incubation as measured by BrdU incorporation (Figure 4.9b/c/d). Cell cycle arrest after 6h incubation was also confirmed by flow cytometric analysis (Table 4.1a). Indeed, apoptosis was also shown to be evident in HL60 cells after incubation with the iron chelators and HU for 6h (Figure 4.7a). Table 4.2 showed that both the chelators and hydroxyurea increased the amount of apoptosis after 6h incubation and that the apoptosing cells came from the S- and Gq/Gi phases of the cell cycle as measured by BrdU incorporation. These findings are all consistent with RR inhibition as a primary mechanism of cell cycle arrest and apoptosis by HU and iron chelators. The iron chelator, CP20 was found to have a more potent anti-proliferative effect in CD34^ cells arresting the cell cycle at concentrations as low as 33|liM IBE (Table 4.4) compared to DFO, which was only able to arrest the cell cycle at concentrations above lOOjiM IBE. These findings are also supported by the findings of others (Hoyes et al., 1992; Cooper et al., 1996). Bidentate chelators such as CP20 and hexadentate chelators such as DFO have been shown to interact differently with human lines, especially in the kinetics of their inhibition of cell cycle (Hoyes et al., 1992; Cooper et al., 1996). Indeed Cooper et al., 1996 using EPR spectroscopy of intact K562 cells showed that the tyrosyl radical of ribonucleotide reductase could be completely removed within 5 minutes incubation with bidentate chelators, but took over 4 hours with the hexadentate chelator, DFO. The authors suggested that this difference could not be explained by the relative cellular uptake rates of the chelators. Instead they hypothesised that the rate of the tyrosyl radical inhibition observed with DFO was compatible with an indirect effect of the drug on incorporation of iron into the apoprotein. By contrast, the rate of inhibition with the bidentate chelators was found to be more compatible with a direct interaction of these ligands with the binuclear ferric iron centre of the enzyme. The findings in this chapter suggest that the consistently higher degree of apoptosis seen with HU compared with the chelators is likely to be due to the different mechanisms of action, namely direct scavenging of the tyrosyl radical by HU and inhibition of iron incorporation by the chelators. These differences may also explain why HU appeared to act more efficiently on the later more mature proliferating cells than the chelators. 162 It is however clear that the disparity between the low proportion of thymocytes in cycle and the high proportion of cells apoptosing with chelators must imply a different mechanism of RR inhibition in thymocytes. In the next chapters, alternative mechanisms for chelator induced thymocyte apoptosis are explored. 163 CHAPTER 5 Involvement of Zinc in Iron Chelator-Induced Thymocyte Apoptosis 164 5.1 Introduction In the previous chapter, the putative role of ribonucleotide reductase and DNA synthesis inhibition as the mechanism for apoptotic induction was explored. These findings suggested that additional or alternative mechanisms are likely to be involved, particularly in thymocytes of chelator-induced apoptosis. In this chapter, the question as to whether the chelation of other metals and in particular zinc, are implicated in the induction of thymocyte apoptosis is explored. There is ample evidence that the concentration of free intracellular divalent metal ions is important in the induction and inhibition of apoptosis in many cell types (Goldstein et a i, 1991 and Orrenius et a i, 1992). Zinc appears to be particularly important in this regard. In humans, the whole body zinc content ranks second among trace elements only superseded by that of iron, there being 2 to 3g of zinc in adult humans. However, no specific biological role for zinc was established until 1940 by Keilin and Mann, when it was shown to be required for the catalytic activity of carbonic anhydrase. In the following 5 decades, the number of known zinc containing enzymes has mushroomed to more than 300. In most of these enzymes, zinc is directly involved in the catalysis, interacting with the substrate molecules undergoing transformation. However in a few enzymes, zinc plays a purely structural role (Berg and Shi, 1996). In contrast there are many fewer different iron metalloproteins or enzymes, and even less numbers have been found to contain copper, molybdenum, selenium, nickel, manganese, or cobalt. There are two properties of zinc which need to be highlighted. The first, unlike other metals, zinc is virtually non toxic (Bertholf , 1988). The homeostatic mechanisms that regulate the entry into, distribution in, and excretion from cells and tissues are so efficient that no disorders are known to be associated with its excessive accumulation, in contrast to iron and other metals. Second, its physical and chemical properties, including its general stable association with macromolecules and its coordination flexibility, make it highly adaptable to meeting the needs of proteins and enzymes that carry out diverse biological functions. These and yet other chemical properties form the basis for the extensive participation of zinc in protein, nucleic acid, carbohydrate, and lipid metabolism, as well as in the control of gene transcription and other fundamental biological processes (Shankar et al., 1998). Zinc has been shown to be an important element for both the development and maintenance of the immune system (Table 5.1) (Kruse-Jarres, 1989; Cunningham- Rundles et at., 1990; Wellinghausen et a l, 1997) and zinc-deficient persons are known to experience an increased susceptibility to a variety of pathogens. It is clear that zinc affects 165 Table 5.1 Effects of zinc on the immune function (Adapted from Wellinghausen et al, 1997) Effects of zinc deficiency Effects of zinc supplementation \,In vivo ^Decreased thymic involution *Reversed thymic involution O) O) * Decreased serum thymulin level *Increased serum thymulin level ^Decreased delayed hypersensitivity ^Impaired immune functions restored ^Decreased peripheral T-cell count *Increased proliferative T-cell response to PH A ^Decreased proliferative T-cell response to PH A ^Increased CD4+ cell count in AIDS patients ^Decreased cytotoxic T-cell activity *Clinical benefit in rheumatoid arthritis * Decreased T helper cell function ^Clinical benefit in the common cold ^Decreased natural killer cell activity ^Decreased macrophage functions 2. In vitro ^Decreased neutrophil functions ^Increased lymphocyte blast transformation ^Decreased antibody production ^Increased neutrophil functions *Increased IFN-ot production in leukocyte cultures ^Increased cytokine production, eg IL-1 in PBMC's ^Increased lymphocyte receptor expression ^Increased LPS-induced cytokine release multiple aspects of the immune system, from the barrier of the skin to gene regulation within lymphocytes. Zinc is crucial for normal development and function of cells mediating non-specific immunity such as neutrophils and natural killer cells. Zinc deficiency also affects development of acquired immunity by preventing both the outgrowth and certain functions of T lymphocytes such as activation, Th 1 cytokine production, and B lymphocyte help (Shankar et a l, 1998). An impaired immune response has been linked to low plasma zinc or to noticeable zinc deficiency in various diseases, particularly in the zinc malabsorption syndrome. Acrodermatitis Enteropathica.. This genetic zinc deficiency syndrome in humans is transmitted as an autosomal recessive trait disease. The disorder is characterised by thymic atrophy and disturbed leucocyte functions associated with recurrent severe infections and is often associated with decreased absorption of ingested zinc. All these immunological defects were found to be reversible by means of zinc supplementation (Neldner er fl/., 1978). The effects of zinc on key immunologic mediators is rooted in the myriad roles of zinc in basic cellular functions such as DNA replication, RNA transcription, cell division and cell activation. Apoptosis is potentiated by zinc deficiency (Shankaret a l, 1998). Labile pools of intracellular zinc are thought to be essential at several points in the mitotic cycle, since a number of enzymes involved in DNA synthesis such as thymidine kinase and DNA polymerase a are affected by zinc deficiency (Chesters, 1989). There is evidence that zinc is required for a critical process in the mid Gi phase of the cell cycle. Under conditions of adequate zinc supply, cells can progress beyond this point to either replicate or differentiate depending on other environmental factors; but if zinc availability is limiting, the cells tend to be diverted into apoptotic cells death (Figure 5.1). Therefore zinc may be involved in regulation of cell numbers by it's role in both proliferation and death by apoptosis. The influence of zinc on apoptosis is well documented and both in vivo and in vitro studies have shown that zinc influences apoptotic cell death ( Frankeret a l, 1977, Martin et a l, 1991, Migliorati et a l, 1993 and Kuo et a l, 1997, Sakabe et a l, 1998; Marini et al., 1998) in particular zinc supplementation in thymocytes is known to inhibit apoptosis induced by agents such as glucocorticoids and gamma irradiation. Although the concentrations of zinc required to prevent apoptosis are high, with most experimental works reporting prevention of DNA fragmentation supplied with zinc concentrations ranging from 250jiM to 5mM (Fukamachi et a l, 1998). These are high concentrations of zinc which are 10-100-fold greater than the concentration of zinc found in serum or tissues, but one which effectively blocks a wide array of cell death inducers for a variety of cell types. Indeed, these observations coupled with the fact thatin vivo zinc deficiency leads to 167 marked thymic atrophy {Acrodermatitis enteropathica) and immunodeficiency point to a regulatory role for zinc in thymocyte apoptosis. Quiescence G2/M G0 /GI Zn Myogenic G1 Differentiation APOPTOSIS Figure 5.1 Relationship of zinc to the cell cycle. Requirement for zinc during the mid Gl phase of the cell cycle and its potential to exit from the cell cycle during differentiation or apoptosis. Current experimental evidence therefore supports four major conclusions:1. zinc- deficiency, resulting from dietary deprivation of mice, or exposure of cultured cells to membrane permeable zinc chelators can induce apoptosis; 2 . zinc supplementation, either by pre-treating mice with zinc sulphate, or adding zinc to the media of cell cultures can prevent apoptotic death. Zinc protects against apoptosis induced by diverse physical, chemical, or immunologic stimuli in cultured cells of lymphoid, hepatic or neoplastic origin; 3. zinc does not affect the triggering events or earliest signs of apoptosis, but acts later in the apoptotic pathway, preventing endonucleosomal fragmentation and subsequent cytolysis. Some authors however, suggest that zinc modulates events in the execution phase of programmed cell death, the execution phase being marked by the sequential activation of the caspases (section 1.7), a family of proteases (Perry et al., 1997; Fukamachi Y et a l, 1998); and 4. an intracellular pool of chelatable zinc plays a critical role in apoptosis, possibly by modulating the activation or activity of the endonuclease. Zinc forms three different complexes with monoprotonated bidentate ligands such as hydroxypyrones, ZnL+, ZnL2 and ZnLg", therefore because the iron chelators, DFO and bidentate HPO's like CP20 have an affinity for zinc as well as iron The affinity constants for the interaction between zinc and 3-hydroxypyridin-4-one are log kl, 7.35; log 168 k2, 6.38; log k3, 5.38; and pKa 9.74 (Taylor et a l, 1988). In principle iron chelators may induce apoptosis by either depriving cells of zinc or by redistributing intracellular zinc pools. The aims of this chapter therefore have been to investigate the interaction of iron chelators with intracellular zinc pools and the relationship of zinc chelation to iron chelator induced apoptosis in thymocytes. 5.2. Effect of iron Chelators on intracellular zinc levels. R ationale Zinc is known to be an important modulator of apoptosis in thymocytes and zinc chelators such as TPEN are known to induce apoptosis in these cells. Furthermore, as outlined in the introduction (section 5.1) both CP20 and DFO have a significant affinity for zinc. Modulation of intracellular zinc pools by iron chelators could in principle initiate apoptosis. It was therefore decided to examine whether clinically used iron chelators could alter intracellular zinc levels and whether apoptosis was affected by such changes. Zinquin was used to determine concentrations of intracellular zinc as previously described (Zalewski et a l, 1993). This method relies on the principle that zinquin contains an ethyl ester which following cleavage by intracellular esterases, carries a negative charge and thereby allows the zinquin to become trapped within the cell. Zinquin fluoresces (wavelength 490nm) when bound to zinc, Zn (II) but not when other metals are bound (Zalewski gr a/., 1993). Preliminary control experiments in thymocytes showed that the addition of iron (in the form of ferric citrate, 300|iM) to solutions of zinquin had no effect on the fluorescence of the free ligand or of the pre-formed Zn-Zinquin complexes. Furthermore the addition of iron chelators (CP20 or DFO), 300p.M IBB had no immediate effect on the fluorescence of pre-formed zinquin-Zn complexes in vitro and therefore changes in levels of free iron within cells modulated by iron chelators would be unlikely to affect zinquin fluorescence. Experimental procedure. Measurement of intracellular zinc concentration in thymocytes 1x10^ thymocytes/ml in culture medium (RPMI medium supplemented with 10% FCS and 2% Pen/Strep) were incubated for 24h in 5ml polypropylene tubes at 37°C and 5% CO2 together with either PBS (as a negative control), CP20 and DFO (300pM IBE) and 169 n Dexamethasone (10" M, as a positive control). Thymocytes were treated for 24h with the drugs prior to treatment for subsequent analysis on the FACS (see methods section 2.2.1). Zinquin, a membrane permanent fluorophore specific for zinc was used with spectrofluorimetry to reveal and quantify labile intracellular zinc. In order to determine whether the iron chelators, CP20 and DFO could modulate intracellular zinc levels and how rapidly this occurred, zinquin was added to thymocytes at a final concentration, 25jiM in Hanks Buffered Salt Solution (HESS) and incubation continued for a further 30 minutes. To obtain a spectrophotometric measurement, the fluorescence of unloaded cells (due to auto fluorescence and light scattering) was subtracted from the readings to derive zinquin dependent fluorescence. In some experiments, washed thymocytes were lysed in the cuvettes by addition of digitonin (SOpM) and the released Zinquin was saturated with 25|aM zinc sulphate, to derive Fmax. Fmin was derived by further addition of IM HCl to quench Zn-dependent fluorescence (Zalewski etal, 1993). Fluorescence was measured at room temperature in a Perkin-Elmer LS 50 luminescence spectrophotometer. Single excitation and emission spectra peaks were observed at wavelengths of 370nm and 490nm respectively. Fluorescence readings by spectrofluorimetry were converted into pmol of Zn/10^ cells using a standard curve(Figure 5.2) derived by titration of increasing amounts of ZnSOq into a solution of 3|iM zinquin, until the fluorescence was equivalent to that obtained with zinquin labelled cells. To show the effect of iron chelators on intracellular zinc concentrations, 1x10^ thymocytes/ml in RPMI-1640 medium were treated for 24h with PBS and DEX (as controls), the iron chelators, CP20 and DFO 300|iM IBE and the zinc chelator, NNNN-tetrakis (2- pyridylmethyl)ethylenediamine (TPEN) (50|iM) prior to treatment with zinquin. R esu lts. 5.2.1 Rate of fall in intracellular zinc with addition of CP20 and DFO Rate of fall in intracellular zinc at 300jdM CP20 and DFO. It can be seen (Figure 5.3) that when 1x10^ thymocytes were pre-loaded with zinquin (25jiM) for 30 minutes before the addition of either CP20 or DFO 300jiM IBE that the addition of CP20 leads to a fall in intracellular zinc levels within Ih of incubation whereas zinc levels take at least Ih to decrease with the addition of DFO at the same concentration. 170 300 4-«<0 c 3 200 0) o c 0 (0 O 1 0 0 o 3 0 001 .01 1 1 1 0 1 0 0 1 0 0 [ZnS04] nM Figure 5.2: Zinquin Standard curve Increasing amounts of zinc sulphate (0-1 pM) was added to HBSS containing 3pM zinquin. Zinquin fluorescence was measured by spectrofluorimetry at 390nm. The data shown are the mean of 3 independent experiments done in triplicate. 171 e 120 L 4-" e e o 1 0 0 * NI L 80 DFO t O 60 CP20 k. 40 0 30 60 90 120 150 180 TIME (mins) Figure 5.3: Kinetics of intracellular zinc deprivation by iron chelators Thymocytes were pre-loaded with zinquin, 25pM for 30 minutes before the addition of 3GGpM IBE CP20 or DFO or control, PBS. Zinquin fluorescence was measured by spectrofluorimetry at 390nm and the results are expressed as a % of control. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 172 Concentration dependence on of rate of fall of intracellular zinc Further experiments over a concentration range of ll-300jj,M IBE were performed (Table 5.2). There was no significant difference in intracellular zinc levels at low concentrations of chelators (135+15.0 and 140±10.0 pmol/10^ cells for CP20 and DFO respectively at concentrations of 33pM IBB). Significant differences between the chelators was seen at 300|iM IBE (20±4.0 and 40+10.0 pmol/10^ cells for CP20 and DFO respectively), p<0 .0 1 . 5.2.2 Effect of pre-incubation of cells with iron chelators in subsequent zinquin fluorescence. Experimental procedure. In order to be sure that the effects observed above were not in some way due to the interaction of chelators with zinquin rather than the intracellular pool of chelatable zinc, further experiments were performed in which thymocytes, 1x 1 0 ^ cells/ml were pre incubated for 24h with chelators before washing the cells three times to remove the chelators before addition of zinquin (25pM). R esu lts Again it is observed (Figure 5.4) that prolonged incubation of thymocytes with either CP20 or DFO reduces levels of intracellular zinc detectable by zinquin, while the addition of dexamethasone which as shown in chapter 3 is a positive control for apoptosis in this cell type, had no effect in decreasing zinquin fluorescence, p<0 .0 0 1 . Values calculated from a standard curve (seesection 2.5} indicate a fall in the intracellular zinc level from 170pmoles/10^ cells in control cells which is in agreement with previous reported levels (Zalewski et a l, 1993) to 20 and 40pmoles/10^ cells in CP20 and DFO treated cells respectively. Discussion. Several investigators have shown that apoptosis is potentiated by zinc deficiency (Shankar et a l, 1998). Greatly increased apoptotic ceU death in tissues of zinc deficient animals and in cells deprived of Znin vitro, as well as inhibition by supplemental Zn of cell death induced by apoptotic, in vitro and in vivo suggest that suppression of apoptosis is a physiological function of Zn (Martin et a l, 1991, Zalewski et al., 1991, 1993). Studies would imply that there is a labile or readily exchangeable pool of intracellular Zn that is important in the regulation of apoptosis. However, many have suggested that this labile 173 Sample Cone. (p,M IBE) Intracellular Zinc Cone, (pmol/10^ cells) Control 165±14.6 CP20 300 17±3.6*** 1 0 0 94±12.6*** 33 135±15.1* 11 160+4.5 DFO 300 38±9.9*** 1 0 0 80±5.8*** 33 140±10.9* 11 145±15.2 T able 5.2: Concentration dependence on the rate of fall of intracellular zinc Thymocytes were incubated with either control or the iron chelators CP20 and DFO at concentrations of 11, 33, 100 or 300piM IBE for 24h. The cells were then washed x3 in PBS prior to the addition of zinquin, 25pM. Zinquin fluorescence was measured by spectrofluorimetry. The data shown are the mean ± SD of 3 independent experiments done in triplicate. *, **, *** denotes significance of p<0.01, p<0.05 and p<0.001 respectively. 174 V) Q) O 200 (O < o 175 (/) 150 o o 125 s Q. 1 0 0 c N 75 50 ■'■54r 25 < M z m iM . 0 Ctrl CP20 HU DFO Figure 5.4: Effect of 24h chelator incubation on intracellular zinc levels Thymocytes were incubated with controls PBS or DEX (lO'^M) and the chelators, CP20 or DFO, 300|aM IBE for 24h at 37°C/5% CO 2. The cells were washed x3 in PBS prior to the addition of zinquin, 25pM. Zinquin fluorescence was measured by spectroBuorimetry. The data shown aie the mean ± SD of 4 independent experiments done in duplicate. 175 pool of Zn only constitutes a small part of the total cellular zinc, since most of the Zn in cells is tightly complexed to metalloenzymes and is not readily exchangeable (Bettger and O'Dell, 1981). Zinquin is a zinc-specific intracellular fluorophore that has been used to detect labile intracellular zinc. It is readily taken up by living cells such as thymocytes and is retained for several hours. It is essentially non-fluorescent until it is complexed with zinc. Zalewski et a i, (1993) investigated the subcellular localisation of the fluorescence using fluorescence microscopy and suggested that the fluorescence was largely restricted to the extranuclear regions. Furthermore, zinquin probably detects only the less tightly bound Zn in cells and this would include free zinc and zinc loosely associated with cellular proteins and lipids. The major pool of cellular zinc which is very tightly bound to the active sites of enzymes and in zinc-finger transcription factors may not be available for interaction with zinquin. Observations in this section indicated that freshly isolated thymocytes had a labile intracellular concentration of 155+22.0 pmol/10^ cells (data not shown), after 24h incubation at 37°C/5% CO 2 , the labile zinc concentration was 170±18.0 pm ol/ 1 0 ^ cells (Figure 5.4). Such values are in agreement with others, indicating that the total zinc (II) content of lymphoid cells falls between values of 150-200 pmol/10^ cells as measured by atomic absorption spectroscopy (Cheeket a i, 1984; Zalewski et a i, 1993). Zalewski et al., 1993 also suggested that there was a good correlation between the average content of labile Zn in the cells and inhibition of DNA fragmentation. The authors showed that a substantial suppression of DNA fragmentation in CLL cells was evident at labile Zn contents of between 50 and 100 pmol/10^ cells. On the other hand, decreases in labile Zn of only a few pmol/10^ cells induced by short term treatment with the zinc chelator TPEN resulted in large increases in DNA fragmentation. In investigating the relationship between intracellular zinc levels and the addition of iron chelators, observations indicated that either short term or long term incubation of thymocytes with the iron chelators, CP20 or DFO lowered the intracellular levels of zinquin detectable zinc. The fall in these levels was found to be more rapid (Figure 5.3) and more pronounced (Figure 5.4) with CP20 than with DFO, a fall which could be correlated to their magnitude of apoptosis in the thymocyte. Furthermore as outlined in previous chapters the rapid fall in intracellular zinc with CP20 in comparison to DFO may be explained by the physiochemical nature of CP20 compared to DFO. As outlined earlier, Zalewski et al. , (1993) suggested that during apoptosis there was an increase in zinquin detectable Zn(II) from intracellular stores or metalloproteins. Iron chelators have previously been shown to cause an increase in thymocyte apoptosis (chapter 3), however in the study presented in this section, the increased apoptosis by the 176 addition of iron chelators was not accompanied by an increase in zinquin detectable zinc (Figure 5.4). It is likely therefore that the iron chelators induce apoptosis and decrease zinquin fluorescence by chelating intracellular zinc. As fluorescent microscopic studies were not performed to investigate the subcellular localisation of zinquin fluorescence, it was difficult to make any predictions on which pool of zinc the iron chelators were likely to affect. Due to the negatively charged nature of zinquin it becomes trapped in the cytosol, not being able to exit the cell via the plasma membrane or to penetrate the nuclear membrane. Therefore in this experiment it is unlikely that the iron chelators affect nuclear Zn, which has been estimated to constitute 25-40% of cellular Zn, but instead chelates the available cytosolic zinc pool loosely bound to proteins and lipids. 5.3 Effect of in vitro addition of zinc on chelator induced apoptosis R ation ale. Having established that iron chelators lower zinquin detectable intracellular zinc levels, which is consistent with the hypothesis that the lowering of intracellular zinc levels may induce apoptosis. Therefore in principle, if zinc depletion is contributing to apoptosis, then the addition of zinc to thymocytes, prior to a challenge with iron chelators should reduce apoptotic induction. In this section therefore, the effects of adding back zinc on apoptotic induction in thymocytes by chelators and DEX was explored. Experimental procedure 1x10^ thymocytes in RPMI medium were pre-incubated with either 50|iM or 200|iM Zn SO4 for 2h at 31^C/5% CO 2 . After this time either PBS, DEX (10"^M) or the iron chelators, CP20 and DFO, 300|iM IBE were added and incubation continued for a further 24h. Samples were treated and analysed by flow cytometry as outlined in sectio n 2.4.1. To assess the potential toxic effect of high concentrations of zinc-sulphate, samples were also analysed for viability as outlined in section 2.3.2. R e su lts. It can be seen in Table 5 .3 that when zinc is added in the form of zinc-sulphate (200|liM), apoptosis was significantly reduced both in control, chelator and DEX treated cells, p<0.001. However the reduction in apoptosis in control cells is small 8 (.6 %) compared with CP20 (53.6%) or DEX treated cells (61.3%). It is interesting to note that using a zinc concentration of 50pM, the amount of apoptosis is increased, p<0.01 177 Zn added PBS CP20 DFO DEX Apoptosis % - 45.4 ± 4.2 77.4 ± 2.4 68.8 ± 8.2 91.6 ± 1.6 + 36.8 ±2.8* 23.8 ± 6.1*** 29.7 ± 4.4*** 30.3 ± 9.4*** Viability loss % - 21.4 ± 2.4 30.1 ± 5.2 31.5 ± 3.2 58.7 ± 13.8 + 48.7 ± 4.5*** 67.2 ± 8.8*** 72.1 ± 6.7*** 90.5 ± 3.3*** Table 5.3: Effect of 200 jaM zinc on thymocyte apoptosis The effects of in vitro addition of zinc sulphate, 200p,M on the apoptosis and loss of viability of murine thymocytes are shown. Freshly isolated murine thymocytes from 10-12 week old male Balb-C mice were incubated for 24h with either control (PBS), CP20 and DFO (300fAM) or Dexamethasone (10‘^M) prior to be analysed for viability loss by fluorescein-di-acetate/ethidium bromide staining or apoptosis by flow cytometry. Results are the mean ± SD of 6 independent experiments. *, **, *** denotes significance values of p<0.01, p<0.05 and p<0.001 respectively using a students T test comparing + or - Zn. (59.8±3.1, 67.0±2.9 and 19.5+1.5% for CP20-treated cells with 0, 50, 200|liM Zn added respectively) (Figure 5.5). Therefore the concentration of zinc is an important factor in the propensity to apoptosis. However, while apoptosis is decreased by the addition of zinc, there is a loss of cell viability as measured by dye exclusion (Table 5.3), indicating a toxic effect to the thymocytes of endogenous zinc. Discussion. Previous reports have demonstrated that the addition of endogenous zinc to thymocyte cultures and other cell types prevents the induction of apoptosis by a variety of agents (Waring et a l, 1990; Chow et a l, 1992; Franker et a l, 1997). Indeed Franker et a l, suggest that a zinc concentration of 500-1000jiM (a concentration that is 10- to 100- fold greater than the concentration of zinc found in serum or tissues) is one which effectively blocks an array of cell death inducers. In this study a concentration of 200)liM was found to be sufficient to suppress apoptosis induced by iron chelators and DEX. The reduced cell viability that was observed after 26h of thymocyte incubation with 200|LiM ZnS04 confirms previous reports that concentrations of zinc as high as l-5mM (Barbieriet a l, 1992) or 250pM (McCabe etal, 1993) may inhibit the features of DNA fragmentation without protecting the thymocytes from cell death. Evidence which in this study also indicates that zinc at pharmacological doses is toxic to thymocytes. The discovery that ZnS04 at concentrations of 50jiM increases apoptosis, confirms previous reports (Provincialiet al, 1994) and represents a finding which correlates with the immuno-regulating action of low zinc concentrations. The data would suggest that in vitro Zn^+ has a dose-dependent antagonistic effect on apoptosis, acting not only as an inhibitor, but also perhaps in playing a role in the physiological intra-thymic cell selection. 5.4 Effect of dietary zinc supplementation on chelator induced apoptosis in vivo. R ationale The above findings showed that the addition of zinc to thymocytes can inhibit apoptosis as measured by DNA fragmentation studies. However because cell viability was adversely affected by the addition of zinc salts (see above) in relatively high concentrations, it was important to investigate whether modulation of zinc levels by a more physiological process of zinc supplementation in the diet could abrogate the thymocyte apoptosis by iron chelators when given in vivo. 179 80 1 CNJ 60 - (D (A w 40 i O Q. O Q. 20 - < 5 0 2 0 0 Zinc Cone uM Figure 5.5: Effect of zinc concentration on CP2()-induced thymocyte apoptosis Thymocytes were incubated with CP20, 3()()pM IBE and either 0, 50 or 200pM zinc- sulphate for 24h at 37"C/5% COi. The cells were spun and fixed in 70% ethanol and stained with PI for apoptosis measurement by Oow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 180 Experimental procedure. Groups of 8 Balb-C mice (aged 6 - 8 weeks at the start of the experiment) were obtained from Harlan OLAC, Essex. The weight of the animals at the commencement of the study was 20-25g. Mice were housed individually in polypropylene cages with stainless steel lids and tap water and diet [rat and mouse no.3 breeding diet (C)] were given ad libitum. The mice were fed on either a diet of high zinc [500mg/kg of RM3 (C) diet] or on a control diet containing normal levels of zinc [35mg/kg RM3 (C)] which was administered for 30 days. After this time the drugs, CP20 and PBS (as a control) were given at a concentration of 2 0 0 mg/kg in 1ml, i.p daily for 60 days while the respective diets were continued. The four groups of mice ( 8 in each group) were as follows: 1. Control diet, PBS’, 2. Zinc diet, PBS’, 3. Control diet, CP20 and 4. Zinc diet, CP20. After 60 days, the mice were sacrificed by cervical dislocation and the thymus dissected as indicated in section 2.3.5. 1x10^ thymocytes/ml in culture medium was then analysed on the FACS (section 2.4.1). Analysis of the intracellular zinc content of each group of mice was also assessed using zinquin incorporation as described in section 2 .5 . R e s u lts . Effects of zinc supplementation and iron chelators on intracellular zinc levels It can be seen (Table 5.4) that mice fed on a high zinc diet had increased levels of intracellular thymocyte zinc (170pmol/10^ cells for PBS) compared to the mice fed on a regular diet (70pmol/10^ cells for PBS) as measured by zinquin incorporation. Mice on a high zinc diet who received daily CP20 however had lower levels of zinc (17pmol/10^ cells) than those receiving PBS only, p<0.001). In mice receiving a regular diet, significantly lower zinc levels were also seen in CP20 treated mice compared with controls (20 and 70pmol/10^ cells for CP20 and PBS treated mice respectively, p<0.001). Cell number was also not affected by diet or treatment. It should be noted in this series of experiments that the level of thymocyte zinc in control animals was lower than in other experiments and the reason for this is not clear. Nonetheless, the effects of iron chelators on zinc levels in zinc loaded and control mice is clear cut. 181 Diet Drug [Zinc] Apoptosis % Viabiiity ioss % ±Zn (pMol/10^ cells) (at 24h) (at 24h) CO ro - Control 20±0.1 7.2±2.9 73.0±6.9 - CP20 17±1.1 36.1±3.1*** 68.0±10.2 + Control 170±1.7 8.6±2.1 69.6±5.7 + CP20 70±0 7*** 10.1±3.8 69.4±8.1 Table 5.4. Effect of zinc supplementation on thymocyte apoptosis The effects of a high zinc diet on in vivo induced apoptosis and on the effect on intracellular zinc levels is shown . Mice were fed on either a high zinc or normal diet for 30 days, then administered either control (PBS) or CP20 (200mg/kg) for 60 days while continuing their respective diets. At the end of this period the animals were sacificied and their thymocytes removed immediately. Intracellular zinc levels were measured by the zinquin method and the proportion of cells undergoing apoptosis estimated by flow cytometry after 24h overnight in control medium at 37°C (RPMI 1640 medium containing L- glutamine and supplemented with 10% PCS). Viability was assessed by fluoroscein-di-acetate/ethidium bromide staining. Results are the ± SD of 8 mice on the same regime. *** denotes a significance of p<0.001 as measured by the students T test. Ejfects o f zinc supplementation on chelator induced apoptosis. At the end of 60 days treatment with CP20 (200mg/kg) iv, the mice were sacrificed and the thymocytes dissected and analysed for apoptosis by flow cytometric analysis as described in section 2.4.1 It can be seen (Table 5.4) that oral zinc supplementation has abrogated the apoptotic effects of CP20 in murine thymocytes reducing the apoptosis from 36% almost back to levels seen in control cells (10.1 %). It is also important to note that whereas in vitro loading of cells with zinc had decreased cell viability by dye exclusion, no such loss of viability was seen in cells loaded with zinc in vivo using a high zinc diet (Table 5.4). Thus dietary supplementation of zinc decreases chelator induced thymocyte apoptosis without affecting cell viability. 5.5 Relative apoptotic effects of zinc and iron chelators R ationale From the above findings, there is evidence to implicate a role of zinc chelation in thymocyte apoptosis by iron chelators; firstly both HPOs and DFO lower intracellular zinc concentrations to levels which have previously been shown to be associated with apoptotic induction by zinc chelators (Zalewski et al., 1993); secondly the addition of zinc both in vitro and in vivo abrogates the induction of apoptosis by iron chelators. In this section, the zinc chelator TPEN, which is known to induce thymocyte apoptosis (Zalewski et al., 1993; Franker et al., 1997; Shumaker et al., 1998) is used as a tool to examine the mechanisms by which HPOs and DFO induce apoptosis. In initial experiments, the effects of TPEN alone are characterised in our system. Subsequently, the interaction of TPEN with HPOs and DFO are examined in order to clarify the effects of zinc and iron chelation on apoptotic induction. 5.5.1 Effect of TPEN alone on thymocyte apoptosis and intracellular zinc le v e ls . Experimental procedure Murine thymocytes, 1x10^ cells/ml in RPMI medium were incubated in the presence of zinc chelator, TPEN at increasing concentrations of O-lOOpM for 24h at 37®C/5% CO 2 , to investigate the effect of concentration on apoptosis. Furthermore to investigate the effect of incubation time on TPEN induced apoptosis, TPEN was incubated at a concentration of 50|iM for between 0-24h at 37®C/5% CO 2 . after the appropriate time intervals, samples were treated for subsequent apoptotic analysis by flow cytometry 183 In order to assess the effect of the zinc chelator, TPEN on intracellular zinc levels, zinquin incorporation as described in section 2.5 was used. TPEN, was incubated in 1x10^ thymocytes at 50jiM for 24h prior to the addition of zinquin 25)liM and fluorometric analysis. R e su lts. Effect of TPEN on apoptosis The studies shown indicate that TPEN, 50|iM caused 70% apoptosis after 24h compared to control as measured by flow cytometry (Figure 5.6a), indeed there was still a significant amount of apoptosis compared to control (PBS) when using concentrations as low as l|iM TPEN (43.3+1.6 compared to 37.2±1.9% for PBS, p<0.01) (Figure 5.6b). TPEN was also able to cause a significant amount of apoptosis compared to control as early as 30 minutes incubation in thymocytes (21.4+4.2% compared to 6 .2+1.4% for PBS), p<0.01. (Figure 5.6c). Effect o f TPEN on intracellular zinc levels. It can be seen in Figure 5.6d that treatment of thymocytes with TPEN for 24h at 50jJ,M significantly quenched fluorescence with zinquin, p<0.001(22.0+4.0 pmol/10^ cells compared to 170+5.0 pmol/10^ cells for control, PBS) (Figure 5.6d) 5.5.2 Effect of Iron loading of thymocytes on apoptotic induction by zinc and iron chelators. R ationale From the experiments in sections 5.2/3/4, it appears that iron chelators can modulate intracellular levels of zinc both in vivo and in vitro and that changes in intracellular zinc broadly parallel their apoptotic effects. In order to clarify whether zinc depletion by iron chelators is directly implicated as an apoptotic mechanism, it was necessary to undertake further experiments where iron was added to thymocytes prior to challenging the cells with chelators. The effects of this pre-loading cells with iron on apoptotic induction by both ‘iron chelators' and ‘zinc chelators’ (TPEN) was then examined. Experimental procedure. Thymocytes at 1x10^ cells/ml were pre-incubated for Ih with ferric citrate 50pM (from a 18mM stock in O.IM HCl), washed x 2 with medium, then either CP20 (150 pM) 184 80 1 60 - ro 'w.2 o 40 - Q. o Q. < 20 - PBS CP20 TPEN Figure 5.6a: Effect of zinc chelation on thymocyte apoptosis Thymocytes were incubated with either control, PBS or CP20, 300pM IBE or the zinc chelator TPEN, 50)iM for 24h at 37‘’C/5% CO^. The cells were spun in a pellet, fixed in 70% ethanol and stained with PI for apoptosis measurement by (low cytometry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. 80 70 - <0 60 - CO CO o Q. 50 - o a . < 40 - .01 1 1 1 0 100 1000 Cone. TPEN pM Figure 5.6b: Effect of TPEN concentration on thymocyte apoptosis Thymocytes were incubated with increasing concentrations of TPEN, 0. l-50)iM for 24h at 37"C/5%C02. The cells were sun and fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. 185 80 - | 9 TPEN 60 - O(A OD. a 40 - PBS < 20 - 0 1 0 20 30 Time Mrs Figure 5.6c: Effect of incubation time on zinc chelation by TPEN Thymocytes were incubated for between 30mins and 24h with TPEN (50)liM) at 37"C/5% CO,. The cells were spun in a pellet and fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. 0) u 200 -1 (O < o 0) o S Q. 100 - C N Ctrl CP20 TPEN Figure 5.6d: Effect of TPEN on zinquin fluorescence Thymocytes were incubated with TPEN, 50|iM for 24h at 37"C/5%C02. The cells were washed x3 prior to the addition of zinquin, 25|iM. Zinquin huorescence was measured by spectrolluorimetry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. 186 or DFO (50|iM) added then TPEN (50jiM) added and incubated for 24h. Samples were then analysed for apoptosis as described previously using flow cytometry(section 2.4.1) or assessed for intracellular zinc levels using zinquin incorporation(section 2.5) Results and discussion. When thymocytes were pre-incubated with ferric citrate prior to the addition of chelator, no significant increase in apoptosis after 24h was observed when compared to control. In contrast when ferric citrate and TPEN, or ferric citrate iron chelator and TPEN, was added to thymocytes there is a significant amount of apoptosis (55.8±3.2% and 68.5±5.3% respectively) compared to control, PBS (39.5±4.1%), P<0.001 (Figure 5 .7 a ). Using zinquin to assess the levels of intracellular zinc, it is shown (Figure 5.7b) that by adding ferric citrate, alone or in combination with an iron chelator, there is no fall in intracellular zinc levels, confirming previous preliminary experiments insection 5.2 which showed that the addition of iron to zinquin had no effect on zinquin fluorescence. However ferric citrate together with TPEN or ferric citrate with chelator and TPEN, caused a significant decrease in the amount of intracellular zinc, p<0.01 (Figure 5.7b). Therefore it is clear from these experiments that by pre-loading thymocytes with iron, the fall in intracellular zinc that was previously observed when similar concentrations of iron chelators were subsequently added to the thymocytes (section 5.2), is abrogated together with the effect on apoptosis. These findings are consistent with the inhibition of iron chelator induced apoptosis, being due to the excess added iron being scavenged by the iron chelator intracellularly in preference to zinc, thereby decreasing intracellular zinc chelation and thus apoptosis. Therefore iron overload could protect against the thymocyte apoptotic induction by iron chelators by indirect inhibition of zinc chelation (due to excess preferentially chelatable iron) rather than inhibition of a primary iron induced apoptotic mechanism. From these experiments, the possibility that iron chelation is the direct mechanism of apoptotic induction by iron chelators cannot be absolutely excluded. However to examine the relative contributions of zinc and iron chelation more closely, further experiments were designed to examine the interaction of zinc and iron chelator combinations on apoptosis. 187 <0 <0 (fl O Q. O CL < Figure 5.7a: Effect of iron addition to chelator-induced thymocyte apoptosis Thymocytes were pre-incubated for 2h with ferric sulphate, lOOpM prior to the addition of the iron chelators, CP20 and DFO, lOOpM IBE or the zinc chelator, TPEN, 50p.M. The cells were spun and fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in triplicate. 188 200 - w 0) o E a. "F 100 - N CP20+Fe TPEN+Fe CP20+Fe+TPEN Figure 5.7b: Effect of addition of iron on zinquin fluorescence Thymocytes were incubated with either PBS, CP20 (300)iM IBE) TPEN (50)iM), Ferric-sulphate (lOOjiM), CP20+Fe, TPEN+Fe or CP20+Fe+TPEN for 24h at 3 7 "C/5 %C0 2 . The cells were washed x3 prior to the addition of zinquin, 25pM. Zinquin fluorescence was measured by spectrofluorimetry. The data shown are the mean ± of 4 independent experiments done in duplicate. 189 5.5.3 Interactions of zinc and iron chelator combinations. R ation ale. In order to examine such potential interactions, experiments were designed to construct a series of isobolograms using the zinc chelator TPEN together with the iron chelators CP20 or DFO. Isobolograms are a simple way of visualising drug interactions and show lines of equivalent effect, i.e. show the relative concentration of pairs of agents that in combination induce the same magnitude of effect (50% of maximal response was chosen). Synergistic interaction is shown by a line of points curving underneath the diagonal zero line to produce a concave-up Isobologram. This type of interaction indicates that the combined effect of two drugs at given concentrations is greater than the sum of each drug individually, in other words by combining two drugs you obtain a better than expected effect. In contrast, a line if points curving above the zero interaction line, to produce a concave-down isobole, indicates that the two drugs antagonise each other and may not suitable to be used in combination therapy. Finally, a line of points along the zero interaction line indicates that the combination is merely additive (Berenbaum, 1989). Experimental design. Murine thymocytes, 1x10^ cells/ml in culture medium (RPMI and L-glutamine supplemented with 10% PCS and 2% Pen/Strep) were incubated for 24h with the iron chelators, CP20 at a range of concentrations from 0-300|iM/ml and DFO, 0-300|iM in the presence of the zinc chelator, NNNN-tetrakis (2-pyridylmethyl)ethylenediamine (TPEN) (0-50jiM/ml) and analysed for apoptosis as previously described by flow cytometry (section 2.4.1). Results and discussion. Table 5.5 showed that the apoptotic effect of CP20 with concentrations of TPEN was antagonistic with high concentrations of TPEN (25-50|iM) and synergistic with low concentrations of TPEN (0.1-5jiM)(Table 5.5 and Figure 5.8). In contrast however when different concentrations of DFO (0-300pM IBE) were added in combination with TPEN (0-50jiM ) no obvious apoptotic synergy or antagonism was apparent (Table5.6). These findings suggest that zinc and iron chelators interact with each other with respect to apoptotic induction and are likely to be acting on a shared pool of metals. Because the previous experiments above(sections 5.5.2) showed that iron loading has little effect on apoptotic induction by TPEN and because TPEN has little affinity for iron, it is likely that the shared metal is zinc rather than iron. The fact that low concentrations of 190 SAMPLE Cone. piM % APOPTOSIS Control 33.9±4.2 CP20 300 71.1±4.3*** 100 55.3±2.1** 10 39.4±2.8 1 33.0+1.6 TPEN 50 66.9+3.4*** 5 46.3±2.4* 0.1 25.9+2.4 TPEN/CP20 50/300 32.0+1.4 50/100 28.5+2.6 50/10 27.3±2.3 50/1 29.4±4.3 TPEN/CP20 5/300 80.7±3.2*** 5/100 81.1±5.3*** 5/10 83.7±3.9*** 5/1 79.7±2.6*** TPEN/CP20 0.1/300 84.9±1.8*** 0.1/100 71.8±5.4*** 0.1/10 76.7±3.4*** 0.1/1 66.6+5.1*** T ab le 5.5: Effect of CP20 and TPEN combinations on apoptosis Thymocytes were incubated with either CP20 and TPEN alone or in combination with each other for 24h. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. *, ** and *** denotes significance of p<0.01, p<0.05 and p<0.001 respectively as measured by the students T test. 191 300 200 u c Oo o CM O. U 100 0 2 3 4 5 6 TPEN Conc. Figure 5.8: Effect of CP20 and TPEN combinations (isobol plot) Thymocytes were incubated with combinations of CP20 (0-300p,M IBE) and TPEN (0-5|iM) for 24h. The cells were washed, fixed in 70% ethanol and stained with PI for apoptosis measurement by flow cytometry. An 80% apoptotic response was used to assess the affect of the drug combination. The results shown are the mean ± SD of 3 independent experiments done in duplicate. 192 SAMPLE Conc. p,M % APOPTOSIS Control 33.9±4.2 DFO 300 49.3±5.1* 1 0 0 41.6+3.2 1 0 40.6±2.2 1 30.5±3.1 TPEN 50 63.1±4.3*** 5 55.4±4.2** 0 . 1 29.4±1.8 TPEN/DFO 50/300 52.0+3.3** 50/100 48.6±2.2* 50/10 54.7±4.3** 50/1 51.2+0.8** TPEN/DFO 5/300 48.5±3.2* 5/100 48.0±4.1* 5/10 47.4±2.8* 5/1 45.3±4.8* TPEN/DFO 0.1/300 36.9±2.2 0 . 1 / 1 0 0 34.0±1.8 0 . 1 / 1 0 32.4±2.8 0 . 1 /1 31.4±3.6 T able 5.6: Effect of DFO and TPEN combinations on apoptosis Thymocytes were incubated with either DFO and TPEN alone or in combination with each other for 24h. The cells were washed, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 3 independent experiments done in duplicate. *, **, *** denotes significance of p<0.01, p<0.05 and p<0.001 respectively as measured by the students T test. 193 CP20 had a synergistic effect on TPEN induced apoptosis suggested therefore that CP20 may be accessing pools of chelatable zinc which are not available to TPEN, this may include zinc in transcription factors. Although TPEN has a higher affinity for zinc, 2.6x10" (Shumaker et a l, 1998) than CP20, because of its larger molecular weight (Shumaker etal., 1998) and hexadentate structure, it is likely that some pools of zinc are less available to TPEN than the smaller hydroxypyridinone molecules like CP20. This contention is supported by the finding that DFO, does not synergise with TPEN. DFO, although having a similar iron binding constant as CP20, has a hexadentate structure and cannot access metal pools which are hidden within protein clefts (Abeysinghe et a l, 1996) as readily available. A further important difference between DFO and CP20 is that DFO being hexadentate will be less likely to donate zinc or iron to other chelators, once a stable complex has formed. Because CP20 can form 2:1 and 1:1 complexes, the tendency to dissociate is greater. This is demonstrated in the spéciation plots (Figure 5.9). Therefore a unifying hypothesis to explain both the apoptotic synergy of CP20 at low concentrations with TPEN and the lack of synergy of DFO with TPEN, is that CP20 is shuttling zinc from zinc pools which are unavailable to TPEN alone onto TPEN. These findings also support the contention in section 5.4 that zinc chelation by CP20 is likely to be a key mechanism of thymocyte apoptotic induction. The finding that DFO cannot synergise with TPEN (presumably because its hexadentate nature precluded shuttling of zinc) does not exclude zinc chelation as a contributing factor to DFO apoptosis 5.6 The effect of Iron and zinc chelators on the activity of the zinc containing enzyme phospholipase C. R a tio n ale The above experiments suggest that HPOs are able to act synergistically with zinc chelators to induce thymocyte apoptosis. The very low concentrations of HPOs required to induce apoptosis in the presence of TPEN, together with consideration of the stability constants and speciatation plots (see above) suggest that HPOs are acting so as to 'shuttle zinc from sites unavailable to TPEN alone onto TPEN. To test whether the predicted shuttling of zinc could be induced by HPOs onto TPEN, a cell free system, using a zinc containing enzyme, was devised to demonstrate this effect directly. Phospholipase C fromBacillus Cereus species contains zinc in the active site of the enzyme (Ottolenghi et a l, 1965). Phospholipase C is an important enzyme involved in cell 194 3 QTQ C Mole Fraction Mole Fraction o o o o o o o CD O )0 O -n srH C c -n (O Ei CJ1 C/2 C -n o en 3 V- •O o CAG S, -n n 00 3 o 3 ro 3 a . H O ~n CD m "U OJ N) + z o II II O — k bo GO T: signalling, it catalyses the reaction where Phosphatidyl Inositol 4,5 Bis-phosphate is reduced to Diacylglycerol and Inositol 1,4,5 Phosphate. The enzyme is responsible for rupturing the bonds between the oxygen and the phosphate. Diacylglycerol is responsible for activating a large family of protein kinase C isoenzymes which catalyse protein phosphorylation reactions. Inositol 1.4.5 phosphate on the other hand interacts with a receptor on the endoplasmic reticulum to release Ca^^" from internal stores. Experimental procedure Phospholipase C was assayed together with /?-nitrophenylphosphoryl-choline (NPPC), based on the observation that phospholipase C hydrolyses NPPC to phosphorylcholine and p-nitrophenol which is chromogenic. (Kurioka, 1968). Phospholipase C from Bacillus cereus species and NPPC were obtained from Sigma Chemical Co. and used without any further purification. Phospholipase C, O.lmg/ml in 60% sorbitol buffered by 0.25M Tris-HCl (pH 7.2) was incubated for 30mins with PBS and DEX (10"^M) (as controls), the iron chelators, CP20 and DFO, 300|aM IBE and TPEN 50)iM, following this time the reaction was started with the addition of 20mM NPPC in 0.25M Tris-HCl (pH 7.2). The assay was carried out at 35^C in a cuvette with a 10-mm optical path length, and the rate of NPPC hydrolysis by phopholipase C was monitored at 10 minute time intervals (0-60mins) by measurement of p-nitrophenol at 410nm. The molar extinction coefficient of p-nitrophenol at 410nm in 60% sorbitol solution buffered by Tris-HCl (pH 7.2) is 1.54x10^ and this was used in the calculation from the obtained optical density to the no. of moles of NPPC hydrolysed. Results and discussion. Figure 5.10 shows the effect of different concentrations of enzyme on the time course for NPPC hydrolysis, all reactions are essentially linear and for subsequent experimentation a concentration of O.lmg/ml enzyme was used. Figure 5.11 shows the effect of addition of PBS (control), the iron chelators CP20 and DFO, 300|aM IBE, the zinc chelator, TPEN, 50|aM and DEX (10"^M) on the enzyme activity after a 30 minute incubation. At high concentrations (300)aM IBE), both the iron chelators, CP20 and DFO were able to inhibit the activity of phospholipase C (CP20 by 80% and DFO by 70%) compared to control at 60mins, similarly TPEN (50jJ.M) inhibited the enzyme by 90%. Both the controls, PBS and DEX had no inhibitory effect on phospholipase C Table 5.7 shows the dose effect of the chelators, CP20 and DFO on the activity of phospholipase C. A range of concentrations of iron chelators were used, l-300|iM IBE to 196 120 M & e E 1 0 0 c •i 80 "O 0.025mg/ml PLC N Uf 60 0.05mg/ml PLC e L 0.1 mg/ml PLC 40 u û. 20 û. 0 20 40 60 80 Incubation time, min Figure 5.10: Effect of concentration on PLC activity Purified phospholipase C (PLC) fromBacillus Cereus species in 60% sorbitol was used at concentrations of O.lmg/ml, 0.05mg/ml and 0.025mg/ml, prior to the addition of 20mM NPPC in 0.25M HCL (pH 7.2). PLC activity was assessed over Ih and measured spectrophotometrically at 410nm. The data shown are the mean ± SD of 3 independent experiments done in duplicate. 197 120 s e 1 0 0 E 80 PBS 4» DEX « 60 DFO 40 CP20 20 TPEN a.U û. 0 0 20 40 60 80 Incubation time (mins) Figure 5.11: Effect of iron chelators on the zinc-containing enzyme, PLC Purified PLC, O.lmg/ml in 60% sorbitol was incubated for 30 minutes with either PBS, DEX (lO^M) as controls or the chelators CP20 and DFO (300|aM IBE) or TPEN, 50^M. After this time, NPPC, 20mM was added and PLC activity assessed over Ih spectrophotometrically at 410nm. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 198 TIME mins SAMPLE 5 10 20 30 40 60 Enzyme(E) + Substrate(S) 11.2±0.6 26.1±1.8 46.5+1.6 61.4+3.6 84.6±3.6 100.8±1.4 E+S+CP20,300nM 03+0.2*** 2.9±0.6*** 93+0.4*** 13.1±1.2*** 15.1±2.2*** 15.3±0.9*** E+S+CP20, lOOuM 0.5+03*** 3.2+0.4*** 7.9+0.6*** 12.5+0.4*** 15.1±0.6*** 15.3±1.8*** E+S+CP20,33nM 2.0±0.5*** 7.4+0.7*** 20.8+0.8*** 36.3+0.8*** 51.4±0.8*** 67.8±1.4*** E+S+DFO, 300nM 1.4+0.6*** 6.6±0.6*** 11.0±1.2*** 17.6+0.8*** 22.6+1.2*** 28.3+1.2*** E+S+DFO, lOOuM 7.4+0.5* 21.1±0.8* 32.5±0.8** 43.5±1.2** 54.1±2.2*** 58.6±3.2*** E+S+DFO, 33fiM 10.0±1.8 20.6+0.6* 36.6±1.4* 56.6+2.8 77.1+0.6* 87.7+4.2* E+S+DFO, 11 [iM 13.3±0.7 25.4±1.4 45.2+2.2 60.6+3.6 80.7±4.2 101.3±1.8 E+S+DFO, IfiM 19.2+0.6 30.3+0.6 46.7+2.4 61.5+3.2 86.3±1.2 114.8+3.6 Table 5.7: Effect of chelator concentration on PLC activity PLC, O.lmg/ml in 60% sorbital was incubated for 30 minutes with increasing concentrations of chelators, CP20 or DFO, 1(*M- 300|iM IBE. After this time, NPPC, 20mM added and PLC activity assessed over Ih spectrophotometricallyt at 420nm. *, ** and *** denotes significance of p<0.01, p<0.05 and (xO.OOl respectively as measured by the students T test. 120 In o 1 0 0 - o E c ■D 80- Control O TPEN 50uM in >. TPEN 0.1 CP20 I mM 2 60- ■o DFO I ajM DFO ljuM/TPEN O.IpM U CP20 IjuM/TPEN IpM Û. 4 0- Cl. Z 2 0 - 0 20 4060 80 Incubation time (mins) Figure 5.12: Effect of chelator combinations on PLC activity Purified PLC, O.lmg/ml in 60% sorbitol was incubated for 30 minutes with either: PBS; TPEN (SO^iM); TPEN (0.1 piM); CP20 or DFO (1|liM) or CP20 and DFO (l^M) in combination with TPEN (0.1 ^iM). After this time, NPPC, 20mM was added and PLC activity assessed over Ih spectrophotometrically at 410nm. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 230 determine if there were any differences between CP20 and DFO on the activity of PLC. CP20 was found to be more effective at inhibiting the activity of PLC compared to DFO, at a concentration of 100|iM IBE CP20 inhibited the enzyme by 85%, in comparison to DFO at this concentration which only inhibited the enzyme by 40%, p 5.7 General discussion Thymocytes readily undergo apoptosis after treatment with various agents, including glucocorticoid hormones, ionising radiation and environmental agents. It is unclear whether initiation of the apoptotic programme by these treatments is mediated by the same signalling pathway that occurs in clonal deletion (section 3.2), indeed several lines of evidence would suggest that modulation of intracellular zinc levels by iron chelators might be an important apoptotic mechanism in thymocytes. It has been shown that the zinc chelator, NNNN-tetrakis (2- pyridylmethyl)ethylenediamine (TPEN) increases apoptosis in peripheral blood lymphocytes (Treveset a l, 1994) and in rat and human thymocytes (McCabe et a l, 1993), a process which appears to be independent of modulation of intracellular calcium. Severe zinc deficiency, as occurs in Acrodermatitis enteropathica, may be associated with thymic atrophy and immunodeficiency (Atherton et a l, 1979; Franker et a l, 1977). Furthermore zinc deficiency has been described in approximately 14% of patients treated with CP20 (LI) (Al-Refaie et a l, 1995) and overwhelming infection together with thymic atrophy has also been described in one patient (Berdoukas et a l, 1993). There is also evidence which points to zinc playing an important role in modulating thymocyte apoptosis. High levels of exogenous zinc block thymocyte apoptosis by glucocorticoid (Sellins et a l, 1987; Telford et a l, 1997) and ionising radiation (Zalewski et a l, 1993; Mathieu et a l, 1996). This has also been demonstrated in other cell types such as hepatocytes (Sun et a l, 1994) and myelogenous cells (Matsubara et a l, 1994) and probably occurs through inhibition of endonucleases, thereby preventing DNA fragmentation (Sunet a l, 1994). Conversely, low 201 concentrations of intracellular zinc have been shown to accelerate apoptosis of lymphoid and myeloid cell lines in vitro (Trubiani et al., 1996; Ning et ah, 1997) as well as human leukaemia lymphocytes, rat splenocytes and rat thymocytes (Jiang et a l, 1995; Telford et a l, 1995; Mathieu et a l, 1996). Indeed, Telford et a l, (1995) suggested that a zinc concentration below SOOjiM provided limited protection against glucocorticoid-induced death. Furthermore there are several cases where zinc failed to block apoptosis. Cyclophosphamide, an alkylating agent that can cross-link DNA at cytotoxic concentrations, readily induces apoptosis in mature human lymphocytes, however, zinc- sulphate failed to suppress apoptosis. The authors suggested that ".... the failure of zinc to provide protection in this system would seem to be more related to its inability to prevent the structural changes in chromatin and DNA created by the cyclophosphamide than to its ability to block endonuclease activity..." (Pette et a l, 1995). Paramanantham et a l , (1997) using in vitro flow cytometric studies on human Chang liver cells showed that addition of zinc caused cell condensation with DNA fragmentation and suggested that zinc could enhance the generation of hydroxyl free radicals, results implying that zinc supplementation could induce features resembling programmed cell death (PCD). (Paramanantham et a l, 1997). Zinc plays an important role in modulating the activities of zinc finger proteins known to be involved in apoptosis such as poly(ADP) ribose polymerases (Riceet a l, 1992) and protein kinase C (Forbes et a l, 1992) and glucocorticoid receptors (Franker et a l, 1997) and there is recent evidence that proteins with zinc finger like motifs may play an important role in modulating apoptosis in immature T-cells (Luiet a l, 1994) and other cells (Birnbaum et a l, 1994). In order to examine the relevance of intracellular zinc chelation to 'iron chelator' induced apoptosis, I sought to answer two questions. Firstly do clinically used chelators reduce intracellular zinc levels significantly following in vitro or in vivo exposure ? and secondly can apoptotic induction by such iron chelators be abrogated by supplementation of zinc in the diet or by the addition of zinc in vitro ? The results of the studies suggested that the answer to both these questions is affirmative for CP20. The addition of this chelator to thymocytes in vitro leads to a fall in intracellular zinc levels within Ih of incubation (Figure 5.3) as measured by zinquin fluorescence Longer term incubation of thymocytes for 24h in vitro with CP20 or DFO results in a 9 fold or 4.5 fold reduction in intracellular zinc levels respectively (Figure 5.4). The reason for the more pronounced and faster rate of fall in intracellular zinc levels with CP20 compared with DFO is likely to be due both the faster rate of access of CP20 to intracellular metal pools than DFO (Hoyes et a l , 1993) and also to the higher affinity for zinc with CP20 (log (Jg = 19.11) compared with DFO (log pi 202 = 11) (Hider et a l, 1990), a result which is also highlighted if the effect of chelator concentration is examined, where low concentrations of CP20 were more efficient at decreasing the intracellular zinc pool than low concentrations of DFO (Table 5.2). It is also likely however that Zn-chelate complexes of DFO are more stable than those of HPOs like CP20 and at low concentrations CP20 may even donate zinc to intracellular proteins (Hider et a l, 1990). Fluctuating levels of CP20 may result in removal of zinc from one locus within a cell with subsequent donation of zinc elsewhere. It is clear from the in vitro experiments that high concentrations of zinc (200|iM) can inhibit the induction of apoptosis by iron chelators but that this is at the cost of a loss of cell viability at 24h (Table 5.3). It is apparent that the ceU viability requires that intracellular concentrations of zinc should be kept within physiological limits, since both over-supplementation and severe depletion of zinc leads to apoptosis (McCabeet al., 1993; Treves et a l, 1994). In order to determine whether zinc could inhibit chelator induced apoptosis without the loss of viability seen following in vitro incubation, further experiments were performed with mice receiving a high zinc diet. Chronic in vivo administration of CP20 resulted in decreased levels of thymocyte zinc particularly in zinc loaded animals (Table 5.4). The results also show that supplementation of the diet with zinc in vivo (Table 5.4) results in reduced apoptotic induction by CP20, without loss of cell viability. These findings taken together suggest that the lowering of intracellular zinc levels by iron chelators and particularly by hydroxypyridinones can induce apoptosis in thymocytes and that by supplementing the diet with zinc, these effects may be abrogated. The exact mechanism(s) by which intracellular zinc depletion by chelators increase apoptosis remain to be determined. Whilst it is arguable that the effects of zinc in vitro are unphysiological and may be acting at the end stage of thymocyte apoptosis by inhibiting endonucleases, such an argument is less persuasive for thein vivo experiments showing inhibition of thymocyte apoptosis following dietary zinc supplementation where chelator induced apoptosis is abrogated without early loss of cell viability (Table 5.4). Whilst many investigators suggest that the protective effect of zinc on apoptosis is attributed to it's inhibition of a Ca^+ and Mg^+-dependent endonuclease, this theory is contested by a number of observations; 1. In L929 cells, zinc inhibited both TNF-a and etoposide induced cytotoxicity prior to the effects on DNA fragmentation (Fady et a l, 1995). 2. Endonuclease functions in the "execution" rather than the regulation phase of apoptosis, therefore inhibition may not prevent cell death. 3. Zinc inhibits the protease responsible for cleavage of lamins in cell free extracts (Takahashi et a l, 1997). 4. In a cell free system, zinc potently inhibits poly-ADP-ribose polymerase, PARP proteolysis induced both by caspase-3 containing apoptotic extract and by purified recombinant caspase-3, therefore 203 identifying caspase-3 as a novel and proximal site of zinc inhibition in the apoptotic pathway (Perry et a l, 1997). However, as glucocorticoid induced apoptosis is also inhibited (albeit less completely) by zinc (Table 5.3), the possibility that direct zinc chelation by DFO and HPOs is not the primary apoptotic mechanism cannot be completely excluded. However, if the chelators are inducing apoptosis by intracellular zinc chelation, the question as to which zinc pool within cells is chelated arises. As with iron, much of the intracellular zinc is unavailable for immediate chelation being tightly complexed to metalloenzymes and is not rapidly exchangeable (Franker et a l, 1977). Using the fluorescent zinc probe, Zinquin, the intracellular labile chelatable pool has been previously quantified under various conditions and found to be 1 0 -2 0 % of the total cellular zinc values (Zalewski et a l, 1993). Because of it's relatively low affinity for xinc, it has been suggested that Zinquin probably detects only the less tightly bound zinc in cells, including free zinc and that loosely associated with cellular proteins and lipids (Zalewski et a l, 1993). It is unlikely that CP20 binds additional zinc pools, as the affinity of hydroxypyridin-4-ones for zinc. Log 83 19.11 is less than that of zinquin (Hider et a l, 1990) To further address the issue of whether the iron chelators are causing apoptosis by zinc chelation. Isobologram experiments using CP20 and the zinc chelator, TPEN show that in combination the two drugs both have an antagonistic and synergistic effect on apoptosis depending on the concentration of TPEN. The effect on apoptosis at low concentrations of TPEN maybe explained by a "shuttling" mechanism whereby CP20 which has a lower molecular weight than TPEN is able to access zinc normally unavailable for chelation, being tightly bound to transcription factors and metallothioneins and transfers this now free zinc to TPEN (Figure 5.8, 5.9). Furthermore experiments using a zinc- containing enzyme, phopholipase C, show that both CP20 and DFO inactivate this enzyme as measured spectrophotometrically(Figure 5.11), but of interest is that phospholipase C can also be inactivated using low concentrations of both CP20 and TPEN in combination (Figure 5.12) which reinforces the suggested shuttling mechanism. The differences observed between the rates of apoptotic induction and the rates of reduction in intracellular zinc by CP20 and DFO may have clinical relevance and is possibly the reason why zinc deficiency is a side effect in a small proportion of patients receiving CP20. Although the lowest concentration at which substantial apoptosis is induced in vitro is similar for both CP20 and DFO following prolonged exposure (see Chapter 3) it is likely that this is of greater potential clinically relevance for HPOs such as CP20 than for DFO. 204 In conclusion the findings in this chapter showed that iron chelators in current clinical use and most notably CP20, induced thymocyte apoptosis both in vivo and in vitro and that this process is modulated by levels of intracellular labile zinc. In the search of orally active chelators with a lower incidence of unwanted side effects, consideration of the relative rates at which novel chelators lower intracellular zinc levels and induce thymocyte apoptosis may be important. 205 CHAPTER 6 Genetic Control of Chelator-Induced Apoptosis in Thymocytes 206 6.1 Introduction The findings in previous chapters suggested that inhibition of DNA synthesis is a contributing mechanism to apoptotic induction by iron chelators in proliferating cells, but is unlikely to be the cause of accelerated apoptosis in thymocytes. Other mechanisms for apoptotic induction by iron chelators in thymocytes therefore require consideration. In Chapter 5, evidence was found that the chelation of zinc may be an important mechanism by which "iron" chelators induce apoptosis in thymocytes. In this chapter, further work is undertaken to elucidate the mechanism(s) by which apoptosis is induced in thymocytes by chelators. This has been achieved by examining the outcome of inhibiting specific mechanisms involved in triggering or effecting apoptosis. The effect of inhibiting these mechanisms has been compared between iron chelators and other known inducers of apoptosis such as glucocorticoids. Because apoptosis is an active process generally requiring the synthesis of new proteins in order to progress, inhibitors of protein synthesis typically inhibit apoptotic induction whatever the initial trigger involved. The effect of inhibitors of RNA and protein synthesis is therefore examined in this chapter to determine whether iron chelators differ from other known inducers of apoptosis in this regard. The trigger mechanisms for thymocyte apoptosis induction by chelators is also examined in this chapter. DNA damage, resulting from a variety of insults, eg ionising radiation, drugs and viruses, can initiate apoptosis. The potential of iron chelators to induce DNA damage as a trigger to apoptosis, possibly by depriving cells of sufficient deoxyribonucleotides for DNA synthesis or repair is examined. As will be explained in greater detail in this chapter, pi53 plays a critical role in checking cells for DNA damage and in allowing apoptosis to progress if DNA damage has occurred. It is known that certain initiators of apoptosis which involve primary DNA damage, such as ionising radiation induction, are abolished in cells where the p53 mechanism is deficient. The induction of apoptosis by iron chelators has therefore been compared in wild type and p53 mutant thymocytes to determine whether primary DNA damage is likely to be a trigger to apoptosis with iron chelators. This has also been examined in murine granulocytes to determine whether granulocytes differ from thymocytes in this regard. Finally, the effect of inhibitors of proteases on thymocyte apoptosis is compared for iron chelators and glucocorticoids in order to gain insight as to likely common or disparate pathways of apoptosis. Proteases are an essential component of the effector arm of the apoptotic machinery (Figure 6.1) and there is evidence for the activation of different protease pathways depending on the initial stimulus and the cell type involved. The effect 207 of inhibitors of a vaiiety of proteases are examined in this chapter including serine, aspartate, neutral and cysteine proteases. The latter family of proteases were first found to be important in regulating and effecting apoptosis in the nematode, Caenorhabditis elegans and are termed ced proteins (see section 1.7). The mammalian homologues of ced are cysteine proteases with aspartate specificity, formerly referred to as ICE (Interleukin 113- Converting enzyme) like proteases, and now called caspases (Alnemri, 1996). All the Ced- 3/caspases characterised to date induce apoptosis when overexpressed and are synthesised as inactive proenzymes, which are then activated following cleavage at specific sites. Use of synthetic oligopeptide inhibitors (as will be investigated in this chapter), designed to mimic the recognition/cleavage sites of known Ced-3/caspase substrates has been useful in clarifying the role of these proteases in apoptosis. extracellular FasL intracellular ,,,, Fas FA D D ^J^^ eath Domain Death Effector Domain C^c^spase- ca sp a se-^ nucleus LAMIN proteolysis PARP cleavage DNA degradation Figure 6.1: Involvement of Caspases in apoptosis 208 6.2 Role of RNA and protein synthesis in chelator-induced thymocyte apoptosis 6.2.1 Rationale Several investigators have shown that transcription and translation of specific genes are required to initiate apoptosis. This assumption is mainly based on the effects of cycloheximide and actinomycin D in preventing thymocyte apoptosis triggered by irradiation, glucocorticoids and Ca^"^ ionophores (Cohen et a l, 1984; Cotter, 1996). However, evidence suggests that the requirement ofde novo protein synthesis during apoptosis is more complex than first thought (Chowet a l, 1995). Previous investigations indicate that there are diverse apoptotic pathways in cells (Cohenet a l, 1984; Martin et a l, 1990; Cohen, 1991). Some apoptotic pathways conform to an ‘induction model’ which requires new RNA/protein synthesis, while other pathways appear to fit a ‘release model’ in which apoptosis is triggered by inhibition of RNA or protein synthesis. In many instances of apoptosis, inhibitors of protein or RNA synthesis have failed to inhibit apoptosis, these include apoptosis of target cells induced by cytotoxic T-lymphocytes (Duke, 1983), apoptosis of macrophages induced by gliotoxin (Waring, 1990) and apoptosis in tumour cell lines induced by mild hyperthermia (Takano et a l, 1991). To compound this conflicting problem, Actinomycin D and Cycloheximide have been shown to induce apoptosis in some normal and neoplastic cell populations (Martinet a l, 1990; Collins et a l, 1991; Bansal et a l, 1990). Indeed Chow et a l, (1995) suggests that inhibition or delay of the apoptotic process by protein synthesis inhibitors could be due to their effects on other cellular processes rather than the inhibition of protein synthesisper se. It is well known that thymocyte apoptosis in many instances can be abrogated by inhibitors of protein synthesis such as cycloheximide (Chow et a l, 1995). Therefore it was the purpose of this study to investigate whether newly synthesised proteins were required for initiation of chelator-induced apoptosis in thymocytes. 6.2.2 Experimental procedure Thymocytes (Section 2.3.5) were incubated with either PBS, the iron chelators CP20 and DFO (300jiM IBE) or DEX (10‘^M) together with either Cycloheximide, 50|Lig/ml (protein synthesis inhibitor) or Actinomycin D, 5pg/ml (RNA transcription inhibitor) for 24h at 37®C, 5% CO 2 The samples were then treated for subsequent quantitation of DNA fragmentation by flow cytometry as outlined in section 2.4.1 209 6.2.3 Results Using cycloheximide as an inhibitor of protein synthesis and actinomycin D as an inhibitor of RNA transcription, the abrogation of iron-chelator induced apoptosis by inhibiting protein and RNA synthesis has been examined. Figure 6.2a shows the effect of cycloheximide on chelator induced apoptosis. In control cells, apoptosis increases from 4.2+1.2% in freshly isolated cells to 38.6+3.3% over a 24h incubation period at 37®C. The addition of cycloheximide significantly decreases the amount of thymocyte apoptosis after 24h in PBS (38.6±3.3% to 16.2+1.8% with the addition of cycloheximide), iron chelator (67.0+3.1% to 12.8+2.9% for CP20 and 54.3±3.8% to 11.9+1.8 % for DFO) and DEX (83.4+4.9% to 10.6+1.6%) treated cells, p<0 . 0 0 1 for all comparisons between treated and untreated thymocytes. Previously, cycloheximide has been shown to protect thymocytes from apoptosis triggered by glucocorticoid hormones, Ca^+ ionophores, or irradiation, but has no apparent effect on DNA fragmentation induced by gliotoxin or the intracellular zinc chelator, TPEN (Aw et a l, 1990; McCabe et al., 1993). The findings in this study suggest however, that newly synthesised mRNA and protein are required for DNA fragmentation and apoptosis in thymocytes exposed to chelators. Figure 6.2b shows the effect of Actinomycin D on chelator induced apoptosis. Actinomycin D has no significant effect on apoptosis in control cells (PBS) (35.4±4.2% to 39.3+4.2% with the addition of Actinomycin D) but significantly inhibited apoptosis induced by the iron chelators CP20 (61.3+3.7% to 39.5+3.1%, p<0.001) and DFO (48.4+2.4% to 40.6+1.9%, p<0.05) or Dexamethasone (DEX) (70.9+5.6% to 33.4+2.8%, p<0.001). This is consistent with DEX and the iron chelators requiring RNA transcription for the induction of thymocyte apoptosis. The finding that Actinomycin D inhibited chelator induced apoptosis, showed that RNA transcription was involved in the apoptotic mechanism of chelator induced thymocyte apoptosis(Fig 6.2b) However as apoptosis is not completely inhibited by inhibitors of RNA transcription, involvement of other mechanisms cannot be ruled out. Indeed, Borrelli et a i, 1992 showed that Actinomycin D inhibited total transcription but did not reduce cytoplasmic levels of rRNA or mRNA. 6.2.4 Discussion. In this study, the role of protein and RNA synthesis following the initiation of thymocyte apoptosis were examined. The results of the studies showed that like many other inducers of apoptosis, iron chelators require newly synthesised proteins using cycloheximide and actinomycin D as inhibitors of protein and RNA synthesis respectively 210 100 - CHX + CHX PBS CP20 DFO DEX Figure 6.2a: Effect of cycloheximide on chelator-induced thymocyte apoptosis Thymocytes were pre-incubated for 2h with cycloheximide (50pg) prior to the addition of 3QGpM IBE chelators or DEX (lO’^M). Control cells received an equal volume of PBS. The cells were washed then incubated at 37°C/5% CO; for a further 24h. The cells were then spun in a pellet, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by Bow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. O » -Act D o 4-Act D o. o o. < PBS CP20 DFO DEX F igure 6.2b: The effect of Actinomycin D on chelator-induced apoptosis Thymocytes were pre-incubated for 2h with Actinomycin D (5pg) prior to the addition of 300pM IBE chelators or DEX (lO^M). Control cells received an equal volume of PBS. The cells were washed then incubated at 37”C/5% CO; for a further 24h. The cells were then spun in a pellet, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 211 (Figures 6.2a & b). Recently, a réévaluation of the role of protein synthesis in the rat thymus however contended the protective effects of cycloheximide (Chow et a l, 1995). The investigator's results suggested that the protective effects on thymocyte DNA fragmentation observed with cycloheximide may only be delaying the onset of chromatin degradation rather than preventing it. Why some cells require the synthesis of proteins in order to undergo apoptosis, when other cells do not is currently not known. One can assume that in cells which do not require new protein synthesis to undergo apoptosis, effector molecules must already be present within the cell and are either regulated by continuous synthesis of specific inhibitors or are compartmentalised within the cell such that they do not cause any damage until specifically regulated to do so. In this study, inhibition of apoptosis by the protein synthesis inhibitor, cycloheximide suggests that iron chelator treatment induces synthesis of a protein(s) required for the cell death process. Alternatively, the requirement of RNA and protein synthesis for the induction of iron chelator induced apoptosis may reflect the need to synthesise molecules that activate existing cell death machinery, rather than making any components required for the cell death program itself. 6.3 Role of p53 in chelator-induced thymocyte apoptosis 6.3.1 Rationale p53 is a sequence specific transcription factor that possesses both transactivational and transrepressive activities. The role of p53 transcriptional activation and damage- induced growth arrest is well understood; p53 induces transcription of the cyclin dependent kinase inhibitor, p 2 1 , which inhibits the exit from G 1 of the cell cycle through its inhibition of Cdk2 and inhibits DNA synthesis by binding to proliferating cell nuclear antigen, preventing its interaction with DNA polymerase a (reviewed by Canman and Kastan, 1995). Early studies indicated that p53-dependent apoptosis occurred in the presence of RNA and protein synthesis inhibitors, suggesting that the mechanism of p53 dependent apoptosis requires functions of p53 other than transcriptional regulation (Caelleset a l, 1994). One of the key roles of p53 is to initiate apoptosis in the presence of DNA damage. The p53 tumour suppressor gene, has been found to be essential for the apoptotic response to DNA damage (Figure 6.3, adapted from DP Lane, 1992). Normal p53 stops cells in the G1 phase of the cell cycle when there is DNA damage. This allows the cell time to repair the damage; however, if this proves impossible, then p53 somehow triggers apoptosis. With mutations in this gene, pivotal repair processes are stymied because of loss 212 Repair before division B DNA Damage p53 levels rise G I arrest or Apoptosis c Division with damage (mutation/aneuploidy) DNA Damage No p53 \ No G 1 arrest Mitotic failure and cell death Figure 6.3: Model for the function of p53 A. Normal cell division, for which p53 is not required, B. In the response of a normal cell to DNA damage, the genome guarding fuction of p53 is induced. C. Cells in which the p53 pathway is inactivated by muation of p53, or by host(MDlV12) or viral oncoproteins, replicate damaged DNA resulting in mutation, aneuploidy, mitotic failure and cell death. Adapted from D.P Lane, 1992. 213 of the p53 function. The result is that the cells with damage to the genes that control proliferation can now progress through the cell cycle outside normal regulatory checks and balances. The end result is tumour development (Cotter, 1996). Thymocytes from p53 knockout mice are resistant to apoptotic death induced by ionising radiation, but apoptosis resulting from exposure to glucocorticoids, or to compounds that may mimic T-cell receptor engagement, is unaffected (Loweet a l, 1993). Since one function of p53 appears to involve the cellular response to DNA damage, these differences may reflect DNA damage as the initiating mechanism following ionising radiation in contrast to glucocorticoids. By comparing chelator-induced apoptosis in various murine cell types, using wild-type and p53 knockout mice, I have sought to determine whether DNA damage by iron chelators is a likely primary event in apoptotic induction. Experimental procedure 6.3.2 p53 status For experiments investigating the effect of p53 on chelator induced apoptosis, homozygote p53 knockout, heterozygote or wild-type C57 Bl/ 6 male mice aged 10-12 weeks were obtained from Professor C. Potten (Patterson Institute, Manchester, UK). These mice were originally obtained by homologous recombination in mice embryonic stem cells to derive a null allele of the gene and this mutated p53 allele was then established into a mouse germ line as previously described (Donehoweret a i, 1992) 6.3.3 Cell type To investigate the effect of p53 on apoptosis , two cell types were investigated: murine thymocytes and granulocytes. Murine thymocytes were dissected from the thymuses of male p53 homozygote, heterozygote or wild-type C57 Bl/ 6 mice aged 10-12 weeks and cultured as indicated in section 2.3.5. The cells were resuspended at a concentration of 1x10^ cells/ml in RPMI 1640 medium containing L-glutamine and supplemented with 10% PCS and 2% Pen/Strep were used for subsequent experimentation. Murine granulocytes (promyelocytes to mature cells) were selected from male p53 homozygote, heterozygote or wild-type C57BL/6 mice aged 10-12 weeks using a rat anti mouse primary antibody (as outlined in section 2.3.6). 1x10^ cells/ml in RPMI-1640 medium containing L-glutamine and supplemented with 10% PCS and 2% Pen/Strep were used for subsequent experimentation. 214 6.3.4 Results and Discussion 6.3.4.1 Effect of P53 on thymocyte apoptosis. Glucocorticoid induced apoptosis is known to be p53 independent whereas apoptosis induced by DNA damage from ionising radiation is p53 dependent{Clarke et at., 1993; Lowe et a l, 1993). If chelators act by inducing primary DNA damage, then some degree of p53 dependence may be demonstrable. In Table 6.1, the effect of incubating thymocytes from homozygous p53 mutants, heterozygous and wild-type animals with PBS (as control), CP20, DFO and DEX (as a positive control) is shown. Thymocytes incubated with either the iron chelators or DEX show an increased amount of apoptosis compared to control in both homozygote (54.7±5.5%, 45.1±6.5% and 75.6±2.6% for CP20, DFO and DEX respectively compared to control, 34.1±4.5%) and normals (52.1±8.1%, 39.6±4.2% and 79.7±2.1% for CP20, DFO and DEX respectively compared to control, 28.1±5.9%). This would suggest that chelator-induced apoptosis is not affected by p53 status and that the mechanism of apoptosis in thymocytes is not through primary DNA damage 6.3.4.2 Effect of p53 in murine granulocytes In Table 6.2, the effect of incubating thymocytes from homozygous p53 mutants, heterozygous and wild-type animals with PBS (as control), CP20, DFO and DEX (as a positive control) is shown. As with the results obtained using murine thymocytes, murine granulocytes from homozygotes, heterozygotes and wild-type animals incubated with either CP20, DFO or DEX no difference was found in the amount of apoptosis with regard to p53 status of the animal. In normal animals as outlined in section 4, only hydroxyurea had a significant effect on the amount of apoptosis compared to control (29.1±2.7% compared to 12.9±1.8%), p<0.001. Using p53 knockout mice, once again only hydroxyurea had an increased effect on the amount of apoptosis (30.8+1.2% compared to 11.1+1.6%), p<0.001. This would suggest that in this cell type as for thymocytes, apoptosis is not dependent on p53 status. p53 is known to play a key role in the regulation of apoptosis in both thymocytes and haemopoietic cells and it has been previously shown that apoptosis induced by DNA damage secondary to ionising irradiation is abrogated in homozygous null p53 thymocytes, while glucocorticoid induced apoptosis is unaffected (Clarkeet al., 1993). The findings in this study show that chelator-induced apoptosis in murine thymocytes and granulocytes was unaffected by knockout of p53 suggesting that, unlike radiation induced apoptosis, chelators may therefore act through a p5 3-independent pathway not involving primary 215 P53 Status PBS CP20 HU DFO DEX + /+ 34.1+4.5 54.7+5.5*** 39.4+4.9 45.1+6.5** 75.6+2.6*** - ! - 28.1+4.3 52.1+8.1*** 33.1+6.8 39.6+3.3* 79.7+2.1*** ro + /- 20.9+2.4 39.8+2.3** 25.6+3.7 33.5+4.7** 81.4+2.8*** O) Table 6.1: Effect of p53 status on chelator-induced thymocyte apoptosis Thymocytes from 12 week old wild type - / homozygote knockout +/+ and heretoxygote (-/+) male mice were freshly isolated and incubated at 37^C with control medium, iron chelators (300 pM IBE), hydroxyurea (ImM) or dexamethasone (10‘^M) for 24h before being analysed for apoptosis by flow cytometry. Results are mean ± SD of 5 experiments done in duplicate with ***, ** and * denoting a significance of p<0.001, p<0.01 and p<0.05 respectively from control mice of the same p53 type. P53 Status PBS CP20 HUDFO DEX +/+ 11.1±1.6 10.8+1.0 30.8+1.2*** 13.8+1.7 8.8+0.1 -/- 12.9+1.8 14.9+1.1 29.1+2.7*** 19.7±2.7 9.2+1.0 ro +/- 10.9+3.4 15.9+2.1 26.4+3.4*** 21.3+2.3 7.3+0.5 Table 6.2: Effect of p53 status on chelator-induced granulocyte apoptosis Granulocytes obtained from murine bone marrow from 12 week old wild type - / homozygote knockout 4-/4- and heretoxygote (-/-f-) male mice were freshly isolated and incubated at 37^C with control medium, iron chelators (300 fiM IBE), hydroxyurea (ImM) or dexamethasone (lO'^M) for 24h before being analysed for apoptosis by flow cytometry. Results are mean ± SD of 5 experiments done in duplicate with *** denoting a significance of P<0.001 from control mice of the same p53 type. DNA damage as a trigger and the pathway through which thymocyte apoptosis is initiated by iron chelators appears to be closer to that found with glucocorticoids than that seen following ionising radiation. 6.4 Role of non caspase-protease inhibitors in chelator-induced thymocyte apoptosis 6.4.1 Rationale Although many of the biochemical events that occur during apoptosis such as externalisation of phosphatidylserine residues in membrane bilayers, selective proteolysis of a subset of cellular proteins and degradation of the DNA into intemucleosomal fragments have been identified (Wyllie et a l, 1984, Lazebnik et ah, 1994 and reviewed by Cryns et a l, 1998), mechanisms by which these events are triggered and regulated are poorly understood. For years many investigators, considered that activation of a specific endonuclease to be the hallmark of apoptosis. Recently however, attention has focused on the activation of proteases, in particular a novel family of proteases related to the Caenorhabditis elegans cell death gene product, ced-3, the so-called caspases (discussed in section 6.5). As well as the caspases, other types of proteases have been implicated as mediators of apoptosis, particularly in the effector mechanisms of apoptosis. Non-caspase proteases shown to have a role in apoptosis include, calpain, cathepsins B and D, collagenase, tissue- type and urokinase-type plasminogen activators (Squieret a i, 1994). Calpain is activated prior to the appearance of the morphological changes of apoptosis and DNA fragmentation in Dexamethasone-treated thymocytes (Squier et a l, 1994). Prior incubation with specific calpain inhibitors (eg. Calpain inhibitors I and H) prevents Dexamethasone and irradiation- induced apoptosis in thymocytes. In vivo substrates for calpain include cytoskeletal proteins, growth factor receptors, transcription factors and signal transducing enzymes. Serine and cysteine proteases are implicated as mediators of apoptosis in a variety of cell types including thymocytes, promyelocytic leukaemic HL60-cells and T lymphocytes. Apoptosis induced in immature rat thymocytes by a variety of inducing agents can be inhibited by N-tosyl-l-lysyl chloromethylketone (TLCK), N-tosyl-l-phenylalanyl- chloromethylketone (TPCK), phenylmethylsulfonyl fluoride (PheMeS02F) and dichloroisocoumarin. TLCK and TPCK inhibit many serine and cysteine proteases. PheMeS02F inhibits serine proteases and a few cysteine proteases and dichloroisocoumarin inhibits a large number of serine proteases. Therefore it would seem clear that more than one protease is required for the apoptotic death of thymocytes. 2:8 This study compares chelator induced apoptosis with Dexamethasone induced apoptosis in murine thymocytes and investigates whether apoptosis could be modulated using the protease inhibitors, leupeptin, pepstatin, aprotinin and calpain inhibitor I. 6.4.2 Experimental procedure To investigate the role of protease inhibitors on chelator-induced apoptosis: 1x10^ thymocytes were pre-incubated for 2 h with various protease inhibitors, aprotinin, 1 mg/ml (Sigma Chemical Co.), leupeptin, 0.5mg/ml (Sigma Chemical Co.), pepstatin, Img/ml (Sigma Chemical Co.) or calpain inhibitor, 25|ig/ml (Boeringer Mannheim, Germany). After this time either PBS, the iron chelators, CP20 or DFO and DEX were added at concentrations indicated in section 2.2. After a further 24h incubation at 37®C/5% CO 2 . Cells were treated for subsequent apoptosis analysis by FACS as outlined in sectio n 2 .4 .1 6.4.3 Results and Discussion 6.4.3.1 Effect of Serine and Aspartic protease inhibitors on Iron chelator induced thymocyte apoptosis In Figure 6.4 the effect of pre-incubating thymocytes for 2h with either aprotinin, leupeptin or pepstatin prior to the addition of chelators is shown. Pepstatin an aspartate protease inhibitor potently inhibits the HIV protease and other aspartic proteases like pepsin, renin, cathepsin D, chymosin and many microbial proteases. In comparison, aprotinin and leupeptin are both serine proteases inhibitors acting on trypsin, plasmin and cathepsins A and B. Aprotinin, leupeptin and pepstatin have no effect on the amount of apoptosis in control cells (38.2±1.3% and 34.3±4.3% and 37.8±1.4% respectively, compared to 35.6±2.7%) (Figure 6.4). In chelator-treated cells the protease inhibitors, aprotinin (Img/ml) and leupeptin (0.5mg/ml) were shown to significantly decrease the amount of apoptosis, p<0.001. In comparison the protease inhibitor, Pepstatin (Img/ml) had no effect on both CP20 and DFO induced apoptosis. In dexamethasone-induced apoptosis only pepstatin decreased the apoptosis (83.0±2.2% to 54.8±5.7% with the addition of pepstatin), p<0 .0 0 1 , the protease inhibitors aprotinin and leupeptin had no inhibitory effect on the amount of dexamethasone-induced apoptosis. Thus while both serine protease inhibitors abrogated CP20 induced apoptosis but not glucocorticoid induced apoptosis, the aspartic protease inhibitor abrogated dexamethasone induced apoptosis with no inhibitory effect on CP20 induced apoptosis. As 219 PBS +Aprotinin +Leupeptin + Pepstatin CP20 +aprotinin +Leupeptin +Pepstatin DEX +Aprotinin +Leupeptin +Pepstatin 20 40 100 % Apoptosis at 24h F igu re 6.4: Effect of Serine and Aspartic protease inhibitors on chelator-induced apoptosis Thymocytes were pre-incubated for 2h with either Aprotinin (Img/ml), Leupeptin (0.5mg/ml) or Pepstatin (Img/ml) prior to the addition of either 300fiM IBE iron chelator or DEX (lO'^M). Control cells received an equal volume of PBS. The cells were washed then incubated at 37°C/5% CO; for a further 24h. The cells were then spun in a pellet, fixed in cold 70% ethanol and stained with PI fcr apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. **denotes a significance of p<0.05 and *** denotes a significance of p<0.001. 220 a full dose response and time course was not performed with each inhibitor, there remains a possibility that these differences are time or concentration dependent. 6,4.3.2 Effect of the neutral protease inhibitory CalpainJ. on chelator- induced apoptosis R ationale Calpain, or calcium-dependent neutral protease, is an enzyme with two isoenzyme forms distinguished by in vitro requirements (Croall and DeMartino, 1991). In vitro substrates for Calpain include cytoskeletal proteins, growth factor receptors, transcription factors and signal transducing enzymes (Murachi et a l, 1989; Wang et a l, 1989). Although a role for calpains in apoptosis has not been firmly established, there is a considerable body of evidence to indicate that this protease is involved in the effector pathways for apoptosis. First, the dysregulation of calcium homeostasis is a prominent feature of many cells undergoing apoptosis, and Ca^"*" fluxes have been shown to result in calpain activation (Wang et a l, 1989). Second, many human diseases in which inappropriate apoptosis is prominent have been linked with calpain over activation, including neurodegenerative disorders and ischaemic brain and cardiovascular damage (Saido et a l, 1994; Sorimachi et a l, 1994 and Lee et a l, 1991). Third, putative substrates for calpain include polypeptides known to be cleaved during apoptosis such as oc-fodrin (Saido et a l, 1993, Martin et a l, 1995). Finally Squier et a l, 1994 showed that pre incubation with calpain inhibitors prevented apoptosis in thymocytes induced by dexamethasone and low level irradiation. Moreover, destruction of the nuclear structural protein lamin B 1, an event that appears to precede endonucleolytic DNA cleavage during glucocorticoid-induced apoptosis in thymocytes is inhibited by a calpain inhibitor (Neamati et a l, 1995). It was therefore of value to know whether calpain inhibitors also inhibited chelator induced thymocyte apoptosis. Results and Discussion In Figure 6.5, the effect of pre-incubating thymocytes for 2h with calpain inhibitor (I) prior to the addition of chelators is shown. In this model, apoptosis in thymocytes induced by dexamethasone was shown to be significantly decreased by the addition of the calpain inhibitor (82.1±3.4% to 43.3±7.1% with the addition of the calpain inhibitor) whereas apoptosis induced by the iron chelators CP20 or DFO was not decreased by the addition of the calpain inhibitor. As with the data presented in section 6.4.1 the fact that no effect was seen with the iron chelators may simply be due to a specific concentration and time dependence of calpain. Regardless, the data does suggest that 221 100 C9 W (O “C31 + C 3 I a. o a < PBS CP20 DFO DEX Figure 6.5: Effect of Calpain inhibitor on chelator induced apoptosis Thymocytes were pre-incubated for 2h with a calpain inhibitor (25pg) prior to the addition of 300pM IBE chelators or DEX (lO'^M). Control cells received an equal volume of PBS. The cells were washed then incubated at 37°C/5% CO; for a further 24h. The cells were then spun in a pellet, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 222 DcaHi Stimulus Adaptor FLIP Active Caspasc Bcl-x Instigators E Active Caspase-V Pj-o-Caspase '1 erniinators XIA P Active Caspase 1 erininators X1A P Death bv Cleuva»e of Key Intracellular Targets Figure 6 .6 : Interacting pathways of apoptosis 223 calpain may only be important for some models of apoptotic cell death but not all and that the induction of apoptosis by iron chelators and dexamethasone may be through different but converging pathways. However, although calpain may be activated by extracellular calcium fluxes during glucocorticoid induced apoptosis, further work is required to assess the exact role of calpain in apoptosis and how it is regulated. 6.5 Effect of cysteine protease (caspase) inhibitors on chelator- induced thymocyte apoptosis 6.5.1 Rationale Caspases consist of a family of at least 10 related cysteine proteases which have been described since the original recognition that CED-3 had sequence similarity with the mammalian cysteine protease interleukin 1 p-converting enzyme (ICE). These proteins are characterised by specificity for aspartic acid in the PI position. AU the caspases contain conserved QACXG (where X is R, Q or G) pentapeptide active-site motif. Studies on a range of experimental systems have suggested that the caspases play a central role in active cell death (reviewed by Hale et al, 1996, Cohen, 1997 and Cryns et a l, 1998). Under normal conditions, caspases are present in the cytosol as inert proenzymes that pose no danger to the integrity of the cell. Several investigators have shown that for a ceU to undergo apoptosis, caspases must be activated by proteolytic processing atAsp sites, an event that leads to a proteolytic cascade among the caspases (reviewed by Crynset a l, 1998) (section 1.7) Two potential interacting, cascade-initiating pathways converge on the activation of downstream effector caspases that act to kill the ceU by cleaving death substrates (Figure 6 .6 ). The first of these pathways is initiated in response to apoptotic stimuli such as DNA damaging agents that trigger the mitochondrial release of cytochrome c into the cytoplasm. In the second pathway, caspase proenzymes are recmited to the ligand bound death receptors via interactions with adapter proteins such as FADD/MORTl, thereby leading to the proteolytic activation of caspases. If chelators induce apoptosis by DNA damage, there may be evidence for involvement of the first pathway. However this would seem unlikely as it was shown in section 6.3 that the chelators induce apoptosis independent of p53 and therefore DNA damage. In this section, the role of caspases in chelator-induced apoptosis in thymocytes is examined and compared with caspase involvement in dexamethasone induced apoptosis by examining the effects of known inhibitors of the caspase system. The effect of two inhibitors of caspases on apoptosis induced by iron chelators, zinc chelators and 224 dexamethasone are compared in order to elucidate the apoptotic pathway involved in chelator induced thymocyte apoptosis. The first caspase inhibitor to be examined is ZVAD-fmk [Benzyloxycarbonyl-Val- Ala-Asp(OMe)fluoromethyl ketone] which has proved to be very useful in a number of previous studies in elucidating the role of caspases in apoptosis (Feamheadet al., 1995; Slee et ah, 1996; McCarthy et al., 1997). ZVAD-fmk is a non-specific broad spectrum cell permeable inhibitor of caspases whose permeability is facilitated by the presence of the benzyl-oxycarbonyl and OMe groups. ZVAD-fmk is a potent inhibitor of apoptosis induced by a wide range of stimuli in a number of different systems, including thymocytes, hepatocytes, human jurkat T cells and neuronal cells.(Chow et al, 1995; Pronk et al, 1996 and Park et al, 1996). ZVAD-fmk inhibits many of the ultrastructural features of apoptosis, including poly (ADP-ribose) polymerase (PARP) cleavage, formation of large fragments and intemucleosomal cleavage (Longthome et al, 1998). The second inhibitor of caspases which has been examined is the inhibitor of Caspase-3 namely, AC-DEVD-CMK. Caspase-3 (YAMA, CPP32, Apopain) is one of the key executioners of apoptosis, being responsible either partially or totally for the proteolytic cleavage of many key proteins such as the nuclear poly-ADP-ribose polymerase (PARP). Caspase-3 is widely distributed, with high expression in cell lines of lymphocytic origin, suggesting that it may be an important mediator of apoptosis in the immune system (Fernandes-Alnemri et al, 1994). The precise role played by caspase-3 in apoptosis is however an area of current confusion, since the production by homologous recombination of mice genetically deficient in caspase-3 has not produced the general failure of cell death seen in Ced-3 deficient nematodes. Caspase-3 deficient mice are smaller than their littermates and die at 1-3 weeks of age and thymocytes from caspase-3 deficient mice show a similar sensitivity to apoptosis induced by a number of different stimuli, including CD95, staurosporine and dexamethasone. Brain development in these deficient mice is however markedly affected, with a variety of hyperplasias being observed from embryonic day 12 (Kuida et al, 1996 and reviewed by Cohen, 1997). 6.5.2 Effect of Caspase inhibitors on chelator induced apoptosis Experimental procedure Thymocytes from male Balb C mice were isolated and cultured as previously described (section 2.3.5). 1x10^ cells/ml were pre-incubated for 2h with the caspase inhibitor ZVAD-fmk (50pM, Bachem, Germany) prior to the addition of either PBS, the iron chelators, CP20 and DFO (300jaM IBB), the zinc chelator, TPEN(50|liM) or 225 Dexamethasone. After a further 24h incubation at 37°C/5% C02 the cells were treated for subsequent analysis by flow cytometry as outlined in section 2.4.1. In further experiments, the effect of AC-DEVD-CMK, a potent cell permeable and irreversible inhibitor of caspase-3 on zinc and iron chelator induced apoptosis was examined. Thymocytes, 1x10^ cells/ml were pre-incubated for 2h with caspase-3 inhibitor, AC-DEVD-CMK (2 0 |llM ) prior to the addition of either, PBS, the iron chelators, CP20 and DFO, the zinc chelator, TPEN (50p.M) or dexamethasone. After a further 24h incubation at 3 7®C/5 % CO 2 , the cells were treated for subsequent analysis by flow cytometry as outlined in section 2.4.1 Results and Discussion In Figure 6.7, the effect of pre-incubating thymocytes for 2h with ZVAD-fmk, prior to the addition of chelators is shown. ZVAD-fmk is a very potent inhibitor of apoptosis induced both by either dexamethasone or by the iron chelators, CP20 or DFO or the zinc chelator, TPEN suggesting that the involvement of caspases is a general phenomenon during apoptosis. The exact role played by ZVAD in inhibiting apoptosis is at present an area of confusion. Recently McCarthy et al., 1997 showed that ZVAD had no effect on initiation of apoptosis, as determined by membrane blebbing, but acts to arrest each apoptotic program before completion. This observation raises the intriguing possibility that membrane blebbing is a discrete subprogram operating during mammalian apoptosis that can lead to cell death via a caspase-independent mechanism. If true this will constrain the potential therapeutic use of caspase inhibitors. In Figure 6.8a, the effect of pre-incubating 1x10^ thymocytes with the caspase-3 inhibitor AC-DEVD-CMK (20pM), for 2h prior to the addition of either PBS, the iron chelators, CP20 and DFO, DEX or the zinc chelator, TPEN for a further 24h is shown. AC-DEVD-CMK had no inhibitory effect on apoptosis in control, DFO and DEX-treated cells. However of interest is that AC-DEVD-CMK significantly inhibited CP20 (60.0+3.1% to 43.2+2.3% with the addition of the inhibitor) and TPEN-induced (67.3±4.0% to 31.0±2.9%) apoptosis (p<0.001). This indicates that involvement of caspase-3 is important for TPEN and CP20-induced apoptosis but not DFO-induced apoptosis at these concentrations. Further experiments were performed to examine the time course of inhibition of apoptosis by AC-DEVD-CMK. In Figure 6.8b the effect of AC-DEVD-CMK, added 2h prior to the addition of CP20 (300|liM IBE) or TPEN (50|iM) and apoptosis is shown at 4, 6 , 8 or 24h incubation with CP20 or TPEN. It can be seen that CP20 induced apoptosis could be significantly inhibited as early as 8 h (39.7+2.1% to 31.6+1.8%), p<0.05 and 226 O (O -ZVAD CO o +ZVAD a o a. < PBS CP20 DFO DEX TPEN Figure 6.7: Effect of the caspase inhibitor, ZVAD on chelator-induced apoptosis Thymocytes were pre-incubated for 2h with ZVAD (50|xM) prior to the addition of either SOOjiM IBE iron chelator, 5 0 |aM zinc chelator, TPEN or DEX (lO'^M). Control cells received an equal volume of PBS. The cells were washed then incubated at 3TC/5% C02 for a further 24h. The cells were then spun in a pellet, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 227 c 100 CN 8 0 - co 01 6 0 - -CSP-3 a Figure 6.8a: Effect of a caspase-3 inhibitor on chelator-induced apoptosis Thymocytes were pre-incubated for 2h with the caspase-3 inhibitor (20pM ) prior to the addition of 300|xM IBE iron chelator, 50pM zinc chelator, TPEN or DEX (lO^M). Control cells received an equal volume of PBS. The cells were washed then incubated at 3TC/5% CO; for a further 24h. The cells were then spun in a pellet, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate SO M 6 0 - W a ■ CP20 ^ +CSP-3 a . 4 0 - a □ TPEN < El +CSP-3 20 - M 0 - 1- 4 6 8 Time (Mrs) Figure 6.8b: Time course of chelator-induced apoptosis and inhibition by a caspase- 3 inhibitor Thymocytes were pre-incubated for 2h with the caspase-3 inhibitor (20pM ) prior to the addition of 300pM IBE iron chelator or 50pM zinc chelator, TPEN The cells were washed then incubated at 37”C/5% CO; for a further 0, 4, 6, 8 or 24h. The cells were then spun in a pellet, fixed in 70% cold ethanol and stained with PI for apoptosis measurement by flow cytometry. The data shown are the mean ± SD of 4 independent experiments done in duplicate. 228 TPEN induced apoptosis could be significantly inhibited as early as 6 h (39.4+3.2% to 29.4±1.6%), p<0.01. 6.6 General Discussion The purpose of this section has been broadly twofold. Firstly to elucidate which particular trigger mechanisms and effectors are involved in chelator induced apoptosis and secondly to determine whether there are differences between "iron" chelators and other mechanisms of apoptosis with respect to involvement of these mechanisms. If for example, the apoptotic pathways for "iron chelators" were similar to TPEN but differed from dexamethasone, this would add support to the idea that zinc chelation is a shared mechanism for TPEN and iron chelators. The results of the studies show that like many other inducers of apoptosis, the induction of apoptosis by iron chelators requires newly synthesised proteins using Cycloheximide and Actinomycin D as inhibitors of protein and RNA synthesis respectively (Figures 6.2a & b. One important protein which results from increased transcription and protein stabilisation is p53, which is responsible for initiating Gl/S-phase cell cycle arrest, DNA repair and/or apoptosis following genotoxic damage. To ascertain the mechanism of chelator induced apoptosis in thymocytes, one possibility was through the iron chelators causing increased transcriptional activation of p53 which mediates gene induction after DNA damage. In this study however it was shown that p53 status does not influence chelator-induced apoptosis (Table 6.1). Other p53 independent stimuli include several that mimic physiological cell deletion signals, namely glucocorticoid, calcium associated activation and ageing in vitro (Clarke et a l, 1993). It would therefore seem possible to assume that iron chelators induce apoptosis through a similar fashion to glucocorticoids. Glucocorticoid induction of apoptosis in the thymocyte is one of the best documented model systems for the study of apoptosis (Cohen et al, 1984). Functional glucocorticoid receptors are required for the induction of cell death by glucocorticoids and there is considerable evidence to support a model of glucocorticoid-induced apoptosis in which glucocorticoid receptor-mediated regulation of gene transcription is a central step (Figure 6.9). 229 Decision to die Glucocorticoid receptor Cell Cycle arrest \ Metabolic changes Transcription Ind action Repression RP-2 c-myc Calmodulin AP-1 activit Receptor d ef iciency Execution Receptor mutation Caspase activationUBcl-2 Lipid peroxidationion — ^ __ . I -"t Calcium fluxes .PARP cleavage Calmodulin Cytosolic calcium elevation inhibitors Intracellular calcium pool depletion H Calpain activation Intracellular Calcium chelatots > Lamin cleavage \ En gulfment Membrane changes Degrad1 ation r / Chromatin condensation High molecular weight DNA fragments Endonuclease activation DNA, RNA, Protein degradation Figure 6.9: Model of apoptosis in glucocorticoid treated lymphocytes: Adapted from Distelhorst, 1997 A number of investigators have postulated that glucocorticoid treatment may trigger apoptosis by enhancing the transcription of a specific 'lysis gene' or 'death gene' (Harrigan et al, 1991). Investigators have sought to confirm the theory that a gene induction event initiates apoptosis by showing that inhibitors of RNA and protein synthesis inhibit apoptosis. Such experiments, however, have produced conflicting information. In rat thymocytes, glucocorticoid-induced apoptosis is blocked by inhibitors of RNA and protein 230 synthesis (Perandones et a i, 1993). In splenic T cells and cell lines representatives of immature T cells, glucocorticoid-induced apoptosis was not inhibited by puromycin and cycloheximide (Perandones et al, 1993). Currently, therefore the actual genes involved in glucocorticoid-induced cell death and the mechanism(s) by which their products signal lymphocyte cell death require further elucidation (Distelhorst CW, 1997). What is known is that the genes that regulate glucocorticoid induced apoptosis are highly conserved from nematode to man (Steller H, 1995). Two genes ced-3 and ced-4 are required for executing cell death, the mammalian homologues of ced-3 are members of a family of caspases. As outlined previously(section 1.7) the first protease recognised in this family was interleukin 1(3 converting enzyme (ICE). However ICE, is not necessary for glucocorticoid-induced cell death (Kuida et al, 1995). Indeed a role for the caspases in executing the death program in glucocorticoid-treated lymphocytes has yet to be investigated in much detail. There is evidence that poly (ADP) ribose polymerase is cleaved during the process of glucocorticoid-induced apoptosis in thymocytes (Kauffmanet a l, 1993). Although the caspases are currently receiving the most attention, other types of proteases have been implicated as mediators of glucocorticoid-induced apoptosis, in particular calpain. In assessing whether the iron chelators induced apoptosis through a pathway similar to that of glucocorticoids, the use of pro tease inhibitors to calpain were used to examine the amount of apoptosis in chelator and glucocorticoid treated cells. The studies showed that there is a potential role for calcium in mediating glucocorticoid-induced apoptosis, but not chelator-induced apoptosis(Figure 6.5). Furthermore iron chelator- induced apoptosis are inhibited by serine protease inhibitors such as aprotinin and leupeptin, but not by aspartic protease inhibitors such as pepstatin. In contrast glucocorticoid-induced apoptosis was inhibited by aspartic protease inhibitors but not by serine protease inhibitors (Figure 6.4). While these differences could be due to time course or concentration effects which have not been studied systematically in this section, taken together with the findings with calpain inhibitors it is likely that dexamethasone induced apoptosis involves a different effector pathway than chelator induced apoptosis. Further work however demonstrated that both chelator-induced and glucocorticoid-induced apoptosis could be inhibited by the caspase inhibitor, ZVAD (Figure 6.7) which suggests that the two inducers of apoptosis may share a common pathway. With respect to inhibition by the general caspase inhibitor ZVAD, there was no qualitative difference between, CP20, DFO, TPEN and dexamethasone, consistent with the idea that caspase are essential for apoptotic induction by all these agents. However, the relatively specific caspase-3 inhibitor, AC-DEVD-CMK, had no effect on dexamethasone 231 or DFO induced apoptosis but inhibited TPEN and CP20 induced apoptosis. This would add support to the idea that CP20 and TPEN have a shared mechanism of apoptotic induction, which is less important with dexamethasone and DFO, possibly Zn chelation, which as highlighted in chapter 5, Zn^^ is a known inhibitor of caspase-3 (Perry et a l, 1997; Pham et al, 1998; D'Amours et al, 1998). 232 CHAPTER 7 Discussion and Conclusions 233 7.0 Conclusions 7.1: Introduction The work presented in this thesis has achieved two principle goals. Firstly it has examined the propensity to iron chelator induced apoptosis in quiescent thymocytes and proliferating HL60 cells and human haemopoietic cells. Secondly, it has investigated the mechanism(s) by which this apoptosis may occur by examining the roles of DNA synthesis inhibition, zinc and the genes and enzymes known to be associated with apoptosis. Iron overload associated with regular blood transfusions is currently treated with the iron chelator, Desferrioxamine (DFO). DFO however, lacks significant oral activity and an orally active chelating agent is urgently required to treat such patients. CP20 is one potentially orally active chelating agent, but suffers from a number of disadvantages which limit its usage. It possesses low kinetic stability, low scavenging power at micromolar concentrations and consequently may form partially dissociated complexes which in principle are capable of generating harmful hydroxyl radicals. A further disadvantage of this prototype compound is that it has been shown to induce apoptosis in a number of different cell types including thymocytes. The purpose of the work presented in this thesis was to provide an understanding of the mechanisms involved in chelator-induced apoptosis. This is important both for the identification of therapeutic agents lacking this apoptotic effect and for the elucidation of the consequences of iron or other metal deprivation on cell proliferation and viability. 7.2: Investigation of chelator-induced apoptosis The data presented in Chapter 3 showed that iron chelators increased apoptosis compared to control in a number of different cell types. These include murine thymocytes and human haemopoietic progenitors (CD34+) in culture after the first 48h and up to 11 days in culture with the growth factors SCF, IL3 and IL6 . Thymocytes are essentially a quiescent (Gq/G i ) cell type and were shown to be susceptible to the effects of iron chelators after a short incubation time of 6 h. In contrast, haemopoietic progenitor cells are also quiescent, but upon stimulation with growth factors, are triggered into the cell cycle and exponentially proliferate. The cells gradually lose their proliferative capacity along with differentiation and arrest in Gq/Gi when released into the peripheral blood as mature cells. In the study, both the iron chelators CP20 and DFO had no apoptotic effect on CD34+ haemopoietic progenitors within 48h of their isolation, when they are known to be predominately in Gq/Gi. Apoptosis was only increased when the cells were exponentially proliferating on day 3-9. 234 Investigation of the propensity of the chelators to cause apoptosis also revealed that CP20 was more potent at lower concentrations than DFO at causing increased apoptosis in thymocytes. As previously hypothesised in this thesis, this could be as a consequence of the physiochemical properties of CP20 in comparison to DFO. By virtue of the low molecular weight and favourable distribution coefficient, CP20 can rapidly penetrate most cells and hence can gain access to a wide range of iron containing proteins, with the potential of inhibiting them. Furthermore, CP20 is also able to mobilise iron from ferritin significantly more rapidly than DFO (Brady et ai, 1988). DFO has a larger molecular weight and a relatively low lipophilicity in the iron complexed form, therefore in this form it does not enter cells rapidly, thereby limiting redistribution of iron in the body. Non-haem iron containing enzymes in which the iron centre is dominated by oxygen and imidazole ligands are particularly susceptible to inhibition by iron chelators such as CP20 and DFO (Hoyes et al., 1992; Abeysinghe et al., 1996; Cooper et al., 1996). Such iron centres are found to be important in enzymes such as ribonucleotide reductase, which is involved in the synthesis of DNA. Indeed CP20 at clinically achievable concentrations was found to inhibit this enzyme and hence DNA synthesis at a much faster rate than DFO (Hoyes et al, 1992; Cooperet al., 1996). As well as inhibiting iron containing enzymes, iron chelators have been shown to induce apoptosis in a number of cells and cell lines (Fukuchi et a l, 1994; Porter et ai, 1994; HïlQti et al, 1995; Ul-Haqerfl/., 1995; Kovtir et al, 1997; Kyriakou et al., 1998). Kovar et al, (1997) found that iron deprivation in the murine lymphoma cell line, 38C13, achieved by either the addition of DFO or the blocking of transferrin receptors with antibodies was sufficient to induce apoptosis. This, the authors attributed to two possible theories. The first is the activation of p53 which acts as a checkpoint control during the G1 phase of the cell cycle. Apoptosis occurringvia p53 activation may result from inhibition of DNA synthesis, caused by insufficient activity of ribonucleotide reductase. Indeed Fukuchi et a l, (1994) showed that DFO was able to cause an increase in p53 expression in several human leukaemic cell types. The second mechanism suggested by Kovar for iron chelator-induced apoptosis involved a decreased in Bcl-2 expression, a protein that is localised in the mitochondrial membrane and has been shown to inhibit most types of apoptotic cell death (Reedet al, 1996; Park et al., 1996). The authors suggested that a decrease in Bcl-2 expression occurs as a result of decreased mitochondrial activity, resulting from impaired function of the iron containing enzymes in the respiratory chain. Kyriakou et al, (1998) suggested that a decreased expression of c- myc levels would be sufficient to induce apoptosis. The increased apoptotic induction in thymocytes by CP20 compared to DFO may have important clinical consequences and may cause the adverse physiological effects observed with CP20 (thymic atrophy). Although no difference between the rates of apoptosis between CP20 and DFO was observed in CD34+ cells, the chelators did 235 increase the amount of apoptosis. It would therefore be clinically important to elucidate the mechanisms involved in chelator-induced apoptosis. 7.3: Mechanisms of chelator-induced apoptosis It is known that iron is essential for growth and viability an that iron deprivation by the addition of iron chelators are known to have anti-proliferative effects in a variety of cell lines and bone marrow progenitors. The anti-proliferative effects are associated with the inhibition of DNA synthesis and inhibition of the iron-containing enzyme, ribonucleotide reductase (Ganeshugara et al., 1980; Lederman et at., 1984; Hoyes et at., 1993). However a causal relationship between the two has yet to be established. One putative mechanism of chelator-induced apoptosis is the inhibition of non-haem iron- containing enzymes involved in DNA synthesis such as ribonucleotide reductase. In chapter 4 the effect of iron chelators on DNA synthesis in various cell types was examined and compared with the effects of hydroxyurea, a known inhibitor of ribonucleotide reductase. The activity of ribonucleotide reductase is cell cycle dependent, increasing markedly at the time of DNA synthesis. Extracts from resting cells contain very low or undetectable activity, whilst there is a dramatic increase in activity as cells enter S- phase. This hypothesis correlates with the results observed in the resting thymocytes where hydroxyurea had little apoptotic effect. The small apoptotic effect with hydroxyurea is entirely consistent with the small proportion of cells in cycle. The chelator, CP20 did however inhibit S-phase cells as measured by ^H uptake experiments, however it was observed that only 5% of the cells in thymocytes are cycling. After 24h incubation, over 50% of the cells were found to be apoptosing after the addition of the chelators, therefore in thymocytes other mechanisms as to the cause of apoptotic induction must be involved, an issue which was examined in chapters 5 and 6. In contrast to the effects of chelators on thymocytes, both the chelators and hydroxyurea were found to have anti-proliferative effects in the HL60 myeloid cell line and in human haemopoietic progenitor cells. In the study presented in this thesis, results suggest that chelator-induced apoptosis in proliferating cell types may be explained by the effects of the chelators on ribonucleotide reductase. In both CD34+ cells in culture and HL60 cells, the results indicated that the cells were susceptible to the effects of hydroxyurea which was shown to broadly parallel the relative sensitivity of the cells to the iron chelators. This is consistent with the idea that ribonucleotide inhibition may be the mechanism of apoptosis in these cells. Furthermore, data from CD34"^ cells in culture showed that cells which were not in cycle (days 0-1) and beyond day 9 were not susceptible to apoptotic induction by iron chelators. BrdU data in the cycling cells which were induced to apoptose by the addition of the iron chelators were shown to come from a population which has recently incorporated BrdU. This means that cells which are making DNA (i.e. in S-phase) are those which are susceptible to apoptotic induction by chelators. DFO was shown to have 236 less of an effect and this is compatible with slower access and the physiochemical properties of DFO in comparison to CP20. This is also compatible with the previously observed differences in the rate of RR inhibition by CP20 and DFO (Cooper et al, 1996). Recently, several authors have suggested other theories to explain the effects of the chelators on proliferating cell types. Both Smithet al, 1995 and Kulp et a l, 1996 showed that iron chelators affected cyclin-dependent kinase activity and DNA synthesis. Furthermore an exciting study by Kyriakou et al, 1998 suggested that a decreased expression of c-myc levels by iron chelators would be sufficient to induce apoptosis as assessed in HL60 cells. Of interest is that another transcription factor which has been shown to have roles in both apoptosis and cellular proliferation is p53, which in the study presented in this thesis was found not to be involved in chelator-induced apoptosis (chapter 6). A more detailed examination of the effects of chelators on genes which encode proteins known to be involved in cell cycle regulation is therefore required. An important consideration in the design of iron chelators is that they should possess a high degree of selectivity for iron (III) in order to minimise long term toxicity. As outlined in chapter 1, a side effect in a small proportion of patients receiving CP20 is zinc deficiency. Apoptosis is potentiated by zinc deficiency and could theoretically be a mechanism of apoptosis by iron chelators. The effect of the iron chelators on zinc was therefore examined in chapter 5. Here it was shown that iron chelators depleted intracellular zinc pools and that apoptotic induction by chelators could be abrogated by zinc supplementation both in vivo and in vitro. Additionally, persuasive data showed that bidentate hydroxypyridinones like CP20 interact with intracellular zinc pools in a fundamentally different manner from the hexadentate iron chelator DFO. Whereas HPOs like CP20 were shown to act synergistically in the presence of the zinc chelator TPEN, enhancing apoptosis at concentrations as low as IpM, DFO has no such effect. This difference is most likely explained by the ability of HPOs to 'shuttle' zinc from sites in the cell onto acceptors of zinc such as metallothioneins. DFO by nature of its hexadentate structure is larger than CP20, and therefore is less able to access some intracellular zinc pools than HPOs. Additionally, metal complexes of DFO are more stable than those of HPOs and is therefore less likely to donate zinc to other binders of zinc. The ability of CP20 to mobilise zinc from intracellular pools was also demonstrated by the work with the zinc containing enzyme, phospholipase C. The clinical implications of these findings are that bidentate chelators may access intracellular zinc pools which are not available to DFO and that if there are physiological acceptors of zinc within the cells, zinc could be shuttled within the cell. The potential of metallothioneins to act as acceptors of zinc from HPOs should be explored in future studies. Also the ability of HPOs to remove zinc from zinc fingers and transcription factors containing such motifs requires systematic investigation. It is possible that this mechanism could be responsible for significant disturbances in cell function. Indeed 237 recent work by Hider (Porter, personal communication) has shown that HPOs can chelate zinc from zinc fingers. A direct link between chelation of zinc from transcription factors and the induction of apoptosis has yet to be demonstrated and requires further investigation. The acceptance that animal cells have a built in death program was a crucial insight into the mechanism of apoptosis. However only when studies in the nematode, caenorhabditis elegans identified genes dedicated to apoptosis and its control; then with the discovery of similar genes with similar functions in the human did the field really expand. The finding that the caspase-3 inhibitor inhibited apoptosis with the zinc chelator TPEN and with the HPO, CP20 but had no effect on DFO or DEX-induced apoptosis is consistent with the concept that zinc chelation may be an important contributory mechanism to apoptosis by HPOs. However the finding that pre-loading thymocytes with iron abrogated both DFO and HPO induced apoptosis suggests that iron chelation is also important to apoptotic induction in thymocytes. The contrasting pattern of abrogation of apoptotic induction with the protease calpain suggests that DEX and CP20 differ in their effector machinery of apoptosis, but that their may be a common final pathway. In conclusion there is evidence provided in this thesis that chelators induce apoptosis in thymocytes by mechanisms which differ from those of cycling cells. In the latter, the findings in this thesis are consistent with the hypothesis that inhibition of ribonucleotide reductase leads to inhibition of DNA synthesis and cell cycle arrest thereby triggering apoptosis. The exact sequence of events which follows after inhibition of ribonucleotide reductase and DNA synthesis leading to apoptosis is not yet clear however. In thymocytes by contrast, additional or alternative mechanisms to inhibition of DNA synthesis must be involved because non-cycling cells apoptose in the presence of chelators. While glucocorticoids are well known to induce apoptosis in these non-cycling thymocytes, the patterns of apoptotic inhibition with protease inhibitors show sufficient differences from those obtained with HPOs to suggest that the pathway for HPOs such as CP20 and DEX induced apoptosis are not contiguous. Recent work suggests that HPOs which are designed so as the interact with intracellular iron pools and non-haem iron containing enzymes at a slower rate induce apoptosis more slowly that the HPOs studied in this thesis (Kayali, 1998). Future work should focus on identifying whether shuttling of zinc within cell compartments occurs in vivo with HPOs and establishing whether the consequence of such shuttling can trigger apoptosis. In the future it would be desirable to design HPOs which are larger and therefore less likely to chelate iron and zinc from relatively inaccessible pools within the cells as well as being more stable in the metal complexed form and thereby able to donate iron and zinc less readily to other sites within the cell. 238 CHAPTER 8 References 239 8.0 References Abeysinghe RD, Roberts PJ, Cooper CE, Maclean KH, Hider RC, Porter JB: The environment of the lipoxygenase iron binding site is explored with novel hydroxypiridinone iron chelators. J. Biol. Chem. 271(14): 7965-7972, 1996 Adamson GM, Harman AW: Oxidative stress in cultured hepatocytes exposed to acetaminophen. Biochem Pharmacol 45:2289-2294,1993 Aglietta M, Piacibello W, Sanavio F et al.: Kinetics of human haemopoietic cells after in vivo administration of granulocytes-macrophage colony-stimulating factor. J.Cli. Invest. 83: 551-557, 1989 Aisen P: The transferrin receptor and the release of iron from transferrin. In Progress in iron research. Hershko C (Ed) New York, Plenum: 31-40, 1994 Ajioka RS, Yu P, Gruen JR: Recombinations defining centromeric and telomeric borders for the hereditary hemochromatosis locus. J Med Genet 34: 28-33, 1997 Alcain FJ, Low H, Crane FL: Iron at the cell surface controls DNA synthesis in CCL39 cells. Biochem. Biophys. Res. Commun. 203(1): 16-21, 1994 Alcain FJ, Low H, Crane FL: Inhibition of DNA synthesis in CCL39 cells by impermeable iron chelators. Biochem. Mol. Biol. Int. 41(2): 303-310, 1997 Alnemri ES, Livingston DJ, Nicholson DW, Salveson G, Thomberry NA, Wong WW, Yuan J: Human ICE/Ced-3 protease nomenclature. Cell. 87:171, 1996 Al-Refaie FN, Hoffbrand AV: Oral iron-chelating therapy, the LI experience. Clin Haematol. 7; 941-964, 1994 Al-Refaie FN, Wilkes S, Wonke B, Hoffbrand AV: The effect of deferiprone (LI) and desferrioxamine on myelopoiesis using the liquid culture system. Br. J. Haematol 87(1): 196-198, 1994 Al-Refaie FN, Hershko C, Hoffbrand AV, Kosaryan M, Olivieri NF, Tondury P, Wonke B: Results of long-term deferiprone (LI) therapy: a report by the international study group on oral iron chelators. Br J Haematol 91: 224-229, 1995 240 Anderson BF, Baker HM, Norris GE: Structure of human lactoferrin; Crystallographic structure analysis and refinement at 2.8A resolution. J Mol Biol 209: 711-734, 1989 Arends MJ, Wyllie AH: Apoptosis: mechanisms and roles in pathology. Int Rev Exp Pathol 32: 223-254, 1991 Armstrong RC, AjaT, Xiang J, Gaur S, Krebs JF, Hoang K, Bai X, Korsmeyer SJ, Karanewsky DS, Fritz LC, Tomaselh KJ: Fas-induced activation of the cell death related protease CPP32 is inhibited by Bcl-2 and by the ICE family protease inhibitors. J.Biol Chem. 271: 16850-16855, 1996 Ashkenazi A, Dixit VM: Death receptors: Signalling and modulation. Science. 281 (5381): 1305-1308, 1998 Atherton DJ, Muller DP, Aggett PJ, Harries JT: A defect in zinc uptake by jejunal biopsies in Acrodermatitis enteropathica. Clin Sci.56:505-507, 1979 Aw TY, Nicotera P, Manzo L, Orrenius S: Tributyl tin stimulates apoptosis in rat thymocytes. Arch-Biochem-Biophys. 283(1): 46-50, 1990 Bailey S, Evans RW, Garratt RC: Molecular structure of serum transferrin at a 3.3A resolution. Biochem 27; 5804-5812, 1988 Bansal N, Houle AG, Melnykovych G: Dexamethasone-induced killing of neoplastic cells of lymphoid derivation: lack of early calcium involvement. J. Cell. Physiol. 143(1): 105-109, 1990 Barankiewics J and Cohen A: Impairment of nucleotide metabolism by iron chelating desferrioxamine. Biochem Pharmacol. 36: 2343-2347, 1987 Barbieri D, Troiano L, Grassilli E, Agnesini C, Cristofalo EA, Monti D, Capri M, Cossarizza A, Franceschi C: Inhibition of apoptosis by zinc; a reappraisal. Biochem. Biophys. Res. Commun. 187(3): 1256-1261, 1992 Bartlett AN, Hoffbrand AV, Kontoghiorghes GJ: Long-term trial with the oral iron chelator l,2-dimethyl-3-hydroxypyridin-4-one (LI). Br J Haematol 76: 301-304, 1990 Becton DL, Bryles P: The anaerobic Esherichia Coli ribonucleotide reductase-subunit structure and iron sulfur center. J.Biol. Chem 48(24): 7189-7192, 1988 241 Berardi AC, Wang A, Levine JD, Lopez P, Scadden DT: Functional isolation and characterisation of human haemopoietic stem cells. Science. 267: 104-108, 1995 Berdoukas V, Bentley P, Frost H, Schnebli HP: Toxicity of oral iron chelator LI. Lancet. 341: 1088, 1993 BerenbaumMC: What is synergy. Pharmacol. Rev. 41(2): 93-141, 1989 Berenson RJ, Bensinger WI, Hill RS, Andrews RG, Garcia-Lopez J, Kalamasz DF, Still BJ, Spitzer G, Buckner CD, Bernstein ID: Engraftment after infusion of CD34+ marrow cells in patients with breast cancer or neuroblastoma. Blood 77(8): 1717-1722, 1991 Berg JM, Shi Y : The galvanization of biology; a growing appreciation for the roles of zinc. Science 271(5252): 1081-1085, 1996 Beris P, Darbellay R, Extermann P: Prevention of p thalassaemia major and Hb Barts hydrops fetalis syndrome. Seminars in haematol 4: 244-261, 1995 Bertholf RL: Zinc. In Handbook on toxicity of inorganic compounds. Seiler HG, Sigel H (Eds). New York Dekker: 788-800, 1988 Bettger WJ, O’Dell BL: A critical physiological role of zinc in the structure and function of biomembranes. Life Sci. 28(13): 1425-1438, 1981 Biemond P, Swaak AJ, Beindorff CM: Superoxide-dependent and independent mechanisms of iron mobilisation from ferritin by xanthine oxidase. Implications for oxygen free radical induced tissue destruction during ischaemia and inflammation. Biochem J 239: 169-173, 1986 Biemond P, Swaak AJ, van Eijk HG: Superoxide dependent iron release from ferritin in inflammatory diseases. Free Radical Biol Med 4:185-198, 1988 Bimbaum MJ, Clem RJ, Miller LK: An apoptosis inhibiting gene from a nuclear polypedrosis virus encoding a polypeptide with CYS/HIS sequence motifs. J. Virol. 68(4): 21-28, 1994 Boldin MP, Varfolomeev EE, Pancer Z, Mett IL, Camonis JH, Wallach D: A novel protein that interacts with the death domain of Fas/APO-1 contains a sequence motif related to the death domain. J. Biol Chem. 270: 7795-7798, 1995 242 Boldin MP, Varfolomeev EE, Pancer Z, Mett IL, Camonis JH, Wallach D: A novel protein that interacts with the death domain of Fas/APO-1 contains a sequence motif related to the death domain. J.Biol. Chem. 270: 7795-7798, 1995 Boldin MP, Goncharov TM, Goltsev YV, Wallach D: Involvement of MACH a novel MGRT-l/FADD-interacting protease, in Fas/APO-1- and TNF receptor-induced cell death. Cell. 85: 803-815. 1996 Borrelli MJ, Stafford DM, Rausch CM, Ofenstein JP, Cosenza SC, Soprano KJ: Cycloheximide protection against Actinomycin D cytotoxicity. J. Cell. Physiol. 153: 507-517, 1992 Bortner CD, Oldenburg NBE, Cidlowski JA: The role of DNA fragmentation in apoptosis. Trends Cell Biol. 5: 21-26, 1995 Bottomley SS: Secondary iron overload disorders. Semin. Haematol. 35(1): 77-86, 1998 Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilising the principle of protein-dye binding. Anal. Biochem. 72: 248-254, 1976 Brady MC, Lilley KS, Treffry, Harrison PM, Hider RC, Taylor PD: Release of iron ferritin molecules and their iron cores by 3-hydroxypiridinone chelators in vitro. J. Inorg. Biochem 35(1): 9-22, 1989 Brandt JE, Bhalla K, Hoffman R: Effects of interleukin-3 and c-kit ligand on the survival of various classes of human haematopoietic progenitor cells. Blood. 83: 1507- 1514, 1994 Brittenham GM: Development of iron chelating agents for clinical use. Blood 80 (3): 569-575, 1992 Brittenham GM, Griffith PM, Nienhuis AW: Efficacy of desferrioxamine in preventing complications of iron overload in patients with thalassaemia major. N Engl J Med 331; 567-573, 1994 243 Bronspiegel-Weintrob N, Olivieri NF, Tyler B: Effect of age at the start of iron chelation therapy on gonadal function in p thalassaemia major. N Engl J Med 323: 713- 719, 1990 Brown DG, Sun XM, Cohen GM; Dexamethasone induced apoptosis involves cleavage of DNA to large fragments prior to intemucleosomal fragmentation. J. Biol. Chem 268(5): 3037-3039, 1993 Burke LC, Bybee A, Thomas NS: The retinoblastoma protein is partially phosphorylated during early G1 in cycling cells but not in G1 cells arrested with alpha- interferon. Oncogene 7(4): 783-788, 1992 Cabantchik ZI, Glickstein H, Mil gram P, Breuer W: A fluorescence assay for assessing chelation of intracellular iron in a membrane model system and in mammalian cells. Anal. Biochem. 233:221-227, 1996 Caelles C, Helmberg A, Karin M: p53-Dependent apoptosis in the absence of transcriptional activation of p53-target genes Canman CE, Kastan MB: Induction of apoptosis by tumour suppressor genes and oncogenes. Semin. Cancer. Biol. 6 : 17-25, 1995 Castriota-Scanderberg A, Sacco M: Agranulocytosis, arthritis and systemic vasculitis in a patient recieving the oral iron chelator Ll( Deferriprone). Brit J Haematol 96: 254- 255, 1997 Cerretti DP, Kozlosky CJ, Mosley B, Nelson N, van-Ness K, Greenstreet TA, March CJ, Kronheim SR, DmckT, Cannizzaro LA, Huebner K, Black RA: Molecular cloning of the interleukin 1 p converting enzume. Science 256: 97-100, 1992 Charlton RW, Jacobs P, Torrance JD: The role of the intestinal mucosa in iron absorption. J. Clin Invest 44: 543-554, 1965 Cheek DB, Hay HJ, Lattanzio L, Ness D, Ludwigsen N, Spargo R: Zinc content of red and white blood cells in aboriginal children. Aus. N. Z. L. Med: 14(5): 638-642, 1984 Chesters JK: Zinc in human biology. Mills CF (Ed) 109-118. Spinger-Verlag London, 1989 244 Chinnaiyan AM, O'Rourke K, Tewari M, Dixit VM: FADD, a novel death-domain containing protein interacts with the death domain of Fas and initiates apoptosis. Cell. 81: 505-512, 1995 Chinnaiyan AM, O'Rourke K, Lane BR, Dixit VM: Interaction of Ced-4 with Ced-3 and Ced-9: a molecular framework for cell death. Science 275: 1122-1126, 1997 Chow SC, Kass GE, McCabe MJ jr., Orrenius S: Tributyltin increases cytosolic free Ca2+ concentration in thymocytes by mobilising intracellular Ca2+, activating a Ca2+ entry pathway and inhibiting Ca2+ efflux. Arch. Biochem. Biophys. 298(1); 143-149, 1992 Chow SC, Peters I, Orrenius S: Reevaulation of the role of de novo protein synthesis in rat thymocyte apoptosis. Exp. Cell Res. 216: 149-159, 1995 Chow SC, Snowden R, Orrenius S, Cohen GM: Susceptibility of different subsets of immature thymocytes to apoptosis. FEBS Lett. 408(2): 141-146, 1997 Civin Cl, Strauss LC, Brovall C, Fackler MJ, Schwartz JF, Shaper JH: Antigenic analysis of hematopoiesis III. A hematopoietic progenitor cell surface antigen defined by a monoclonal antibody raised against KG-la cells. J. immunol. 133; 157, 1984 Clarke AR, Purdie CA, Harrison DJ, Morris RG, Bird CC, Hooper ML, Wyllie AH: Thymocyte apoptosis induced by p53-dependent and independent pathways. Nature.362: 849-852, 1993 Clem RJ, Fechheimer M, Miller LK: Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science. 254(5036): 1388-1390, 1991 Cohen A: Current status of iron chelation therapy with desferrioxamine. Semin Haematol 27: 86-90, 1990 Cohen JJ, Duke RC, Fadok VA, Sellins KS: Apoptosis and programmed cell death in immunity. Annu Rev Immunol. 10: 267-293, 1992 Cohen J; Glucocorticoid-induced apoptosis in the thymus. J. Immunol 4(6); 363-369, 1992 245 Cohen GM, Sun XM, Feamhead H, Macfarlane M, Brown DG, Snowden RT, Dinsdale D: Formation of large molecular weight fragments of DNA is a key committed step of apoptosis in thymocytes. J. Immunol. 153(2): 507-516, 1994 Cohen GM: Caspases: Executioners of apoptosis. Biochem J. 326(pt.l): 1-16, 1997 Collins AF, Dias GC, Haddad S, Talbot R, Herst R, Tyler H, Zuber B, Blanchette VS, Olivieri NF: Evaluation of a new transfusion preparation VS washed cell transfusion in patients with homozygous beta thalassaemia. Transfusion.34: 517, 1994 Collins RJ, Harmon BR, Souvlis T, Pope JH, Kerr JF: Effects of cycloheximide on B- chronic lymphocytic leukaemic and normal lymphocytes in vitro: Induction of apoptosis. Br. J. Cancer. 64(3): 518-522, 1991 Collins S, Gallo R, Gallagher R: Continuous growth and differentation of human myeloid leukaemic cells in suspension culture. Nature. 270: 347, 1977 Collins S, Ruscetti F, Gallagher R, Gallo R: Terminal differentaition of human promyelocytic leukaemia cells induced by dimethylsulfoxide and other polar compouns. Proc Natl Acad Sci USA. 75: 2458, 1978 Collins SJ: The HL60 Promyelocytic leukaemia cell line:proliferation, differentiation, and cellular oncogene expression. Blood. 70: 1233-1244, 1987 Conrad ME: Iron absorption, in Johnson LR (ed): Physiology of the gastointestinal tract. Raven, New York, NY: 1437-1453, 1987 Conrad ME, Umbreit JN, Moore EG: A role for mucin in the absorption of inorganic iron and other metal cations: A study in rats. Gastroenterology 104: 1700-1704, 1993 Conrad ME, Umbreit JN, Moore EG et al.: Alternate iron transport pathways: Mobilferrin and integiin in K562 cells. J. Biol. Chem. 269; 7169-7173, 1994 Cooper CE, Lynagh GR, Hoyes KP, Hider RC, Cammack R, Porter JB: The relationship of intracellular iron chelation to the inhibition and regeneration of human ribonucleotide reductase. J. Biol Chem 271(34): 20291-20299, 1996 Cotter TG, Fernandes RS, Verhaegen S, McCarthy JV: Cell death in the myeloid lineage. Immunol Rev 142: 93-112, 1994 246 Cotter T: The genes and enzymes of cell suicide. The biochemist. June/July, 1996 Cowan JA: Inorganic Biochemistry, an introduction. VCH, Weinheim, 1993 Crichton RR and Ward RJ: Iron metabolism-New perspectives in view. Biochemistry 31: 11255-11264, 1992 Crichton RR, Ward RJ; Iron species in iron homeostasis and toxicity. Analyst. 120(3); 693-697, 1995 Crichton RR; Ward RJ: Iron homeostasis. Met-ions-Biol-Syst. 35:633-665, 1998 Crichton RR: Inorganic biochemistry of iron metabolism, Ellis Horwood Ltd, 1991 Croall DE, DeMartino GN: Calcium-activated neutral protease (calpain) system: Structure, function and regulation. Physiol Rev. 71(3): 813-847, 1991 Cryns VL, Bergeron L, Zhu H, Li H, Yuan J: Specific cleavage of alpha fodrin during Fas and tumour necrosis factor-induced apoptosis is mediated by an interleukin 1 beta converting enzyme/ced-3 protease distinct from the poly(ADP-ribose) polymerase protease. J.Biol.Chem. 271(49): 31277-31282, 1996 Cryns VL, Yuan J: Proteases to die for. Genes and development 12: 1551-1570, 1998 Cunningham JM, Al-Refaie, Hunter AE, Sheppard L, Hoffbrand AV; Differential toxicity of a-keto hydroxypyridine iron chelators and desferrioxamine to human haemopoietic precursors in vitro. Euro J. Haematol 52: 176-179, 1994 Cunningham-Rundles S, Bockman RS, Lin A, Giardina PV, Hilgartner MW, Caldwell-Brown D, Carter DM. Ann Ny Acad Sci. 587: 113, 1990 D'Amours D, Germain M, Orth K, Dixit VM, Poirier GG: Proteolysis of poly(ADP- ribose) polymerase by caspase-3: Kinetics of cleavage of mono(ADP-ribosyl)lated and DNA bound substrates. Radiat. Res. 150(1): 3-10, 1998 Darzynkiewicz Z, SharlessT, Staino-Coico L, Melamed MR: Subcompartments of the G1 phase of the cell cycle detected by flow cytometry. Proc. Natl. Acad. Sci. USA. 77: 6696-6699, 1980 247 De Virgiliis S, Congia M, Frau F: Desferrioxamine-induced growth retardation in patients with thalassaemia major. J Pediatr 113: 661-669, 1988 Dezza L, Cazzola M, Danova M, Carb-Stella C, Bergamaschi G, Brugnatelli S, Invemizi R, Mazzini G, Riccardi A, Ascari E: Effects of desferrioxamine on normal and leukaemic human haemopoietic cell growth, in vitro and in vivo studies. Leukaemia. 3:104, 1987 Distelhorst CW: Glucocorticoid-induced apoptosis. Adv. Pharmacol. 41: 247-270, 1997 Djeha A, Perez-Arellano JL, Brock JH: Transferrin synthesis by mouse lymph nodes and peritoneal macrophages: iron content and effect on lymphocyte proliferation. Blood 81: 1046-1050, 1993 Dobbin PS, Hider RC: Iron chelation therapy. Chem Britain: 565-568, 1990 Dobbin PS, Hider RC, Hall AD, Taylor PD, Sarpong P, Porter JB, Xiao G, Van der Helm D; Synthesis, physiochemical properties and biological evaluation of N- substituted 2-alkyl -3 -hydroxy-4( 1 H)-pyridinones: orally active iron chelators with clinical potential. J. Med. Chem 36; 2448-2458, 1993 Donahue RE, Kirby MR, Metzer ME, Agricola BA, Seller SE, Cullis HM: Peripheral blood CD34+ cells differ from bone marrow CD34+ cells in Thy-1 expression and cell cycle status in nonhuman primates mobilised or not mobilised with granulocyte colony stimulating factor and/or stem cell factor. Blood 87(4): 1644-1653, 1996 Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery CA, Butel JS, Bradley A: Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature. 356: 215-221, 1992 Duan H, Dixit VM: RAIDD is a new 'death' adaptor molecule. Nature. 385: 86-89, 1997 Duke RC, Chervenak R, Cohen JJ: Endogenous endonuclease induced DNA fragmentation: an early event in cell mediated cytolysis. Proc. Natl. Acad. Sci. USA. 80; 6361-6365, 1983 248 Ellis RE, Yuan JY, Horvitz HR: Mechanisms and functions of cell death. Annu Rev Cell Biol. 7: 663-698, 1991 Emerson SF: The stem cell model of hematopoiesis. In Hematology: Basic principles and procedures. Edited by Hoffman R, Benz EJ, Shattil SJ, Furie B, Cohen HJ. New York, Churchill Livingston, 1995 Enari MH, Sakahira H, Yokoyama K, Okawa A, Iwamatsu, Nagata S: A caspase activated DNase that degrades DNA during apoptosis and its inhibitor ICAD. Nature. 380: 723-726, 1998 Epsztejn S, Kakhlon O, Glickstein H, Breuer W, Cabantchik I: Fluoresence analysis of the labile iron pool of mammalian cells. Anal. Biochem. 248(1): 31-40, 1997 Espinosa De Los Monteros A, Sawaya BE, Guillou F, Zakin MM, De Vellis J, Schaeffer E: Brain specific expression of the human transferrin gene. Similar elements govern transcription in oligodendrocytes and in a neuronal cell line J Biol Chem 269: 24504-24510, 1994 Estrov Z, Taw a A, Wang X, Dube I, Sulh H, Cohen A, Gelfand E, Freedman MH: In vitro and In vivo effects of desferrioxamine in neonatal acute leukanaemia. Blood.69: 757-761, 1987 Estrov Z, Cohen A, Gelfund EW, Freedman MH: Synergistic antiproliferative effects on HL60 cells; desferrioxamine enhances cytosine arabinoside, methotrexate and daunirubicin cytotoxicity. Am.J. Pediatr. Haematol. Oncol. 10(4): 288-291, 1988 Evan GI, Wyllie AH, Gilbert CS, Littlewood TD, Land H, Brooks M, Waters CM, Penn LZ, Hancock DC: Induction of apoptosis in fibroblasts by c-myc protein. Cell 69(1): 119-128, 1992 Evan GI, Brown L, Whyte M, Harrington E: Apoptosis and the cell cycle. Curr. Opin. Cell. Biol 7(6): 825-834, 1995 Fady C, Gardner A, Jacoby F, Briskin K, Tu Y, Schmid I, Lichtenstein A: A typical apoptotic cell death induced in L929 targets by exposure to tumour necrosis factor. J. Interferon. Cytokin. Res. 15(1): 71-80, 1995 249 Fadok VA, VoelkerDR, Campbell PA, Cohen JJ, Bratton DL, Henson PM: Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers specific recognition and removal by macrophages. J. Inununol. 148: 2207-2216, 1992 Feamhead HO, Chwalinski M, Snowden RT, Omerod MG, Cohen GM: Dexamethasone and etoposide induce apoptosis in rat thymocytes from different phases of the cell cycle. Biochem. Pharmacol. 48(6): 1073-1079, 1994 Feamhead HO, Dinsdale D, Cohen GM: An interleukin 1-beta converting enzyme-like protease is a common mediator of apoptosis in thymocytes. FEBS-Lett. 375(3): 283- 288, 1995 Feder JN, Gnirke A, Thomas W: A novel MHC class 1-like gene is mutated in patients with hereditary hemochromatosis. Nat Genet 13: 399-408, 1996 Fernandes AT, Amstrong RC, Krebs J: In vitro activation of CPP32 and Mch3 by Mch4, A novel human apoptotic cysteine protease containing two FADD-like domains. Proc Natl Acad Sci USA. 93: 7464-7469, 1996 Fernandes-Alnemri T, Litwack G, Alnemri ES: CPP32, a novel human apoptotic protein with homology to Caenorhabditis Elegans cell death protein Ced-3 and mammalian interleukin 1 p converting enzyme. J.Biol Chem 269:30761-30764, 1994 Fibach E, Sachs L: Control of normal differentiation of myeloid leukaemic cells XI. Induction of a specific requirement for all cell viability and growth during the differentiation of myeloid leukaemic cells. J, Cell. Physiol 89(2): 259-266, 1976 Finch CA, Ragan HA, Dyer lA, Cook JD: Body iron loss in animals. Proc. Soc. Exp. Biol Med 159: 335-338, 1978 Fischer DE: Apoptosis in cancer therapy: Crossing the threshold. Cell 78; 539-542, 1994 Fleming MD, Trenor CC III, Su MA: Microlytic anaemia mice have a mutation in Nramp 2, a candidate iron transporter gene. Nature Genetics 16: 383-386, 1997 Fleming MD, Romano MA, Su MA, Garrick LM, Garrick, MD, Andrews NC: Nramp 2 is mutated in the anemic Belgrade (b) rat Evidence of a role for Nramp 2 in endosomal iron transport. Proc Natl Acad Sci USA 95: 1148-1153, 1998 250 Fontecave M. Ribonucleotide reductases and radical reactions. Cell. Mol. Life. Sci. 54(7): 684-695, 1998 Forbes IJ, Zalewski PD, Giaakis C, Cowland PA: Induction of apoptosis in chronic lymphocytic leukaemia cells and its prevention by phorbol ester. Exp Cell Res. 198(2): 367-372, 1992 Fosburg MT, Nathan DO: Treatment of Cooleys anaemia. Blood 76: 435-444, 1990 Franker PH, Haas SM, Luecke RW: Effect of zinc deficiency on the immune response of the young adult A/J mouse. J Nutrit. 107; 1889-1895, 1977 Franker PJ, Telford WG: A reappraisal of the role of zinc in life and death decisions of cells. Exp Biol Med. 215: 229-236, 1997 Freedman MH, Boy den M, Taylor M: Neurotoxicity associated with desferrioxamine therapy. Toxicology 49; 283-290, 1988 Fukamachi Y, Karasaki Y, Sugiura T, Itoh H, Abe T, Yamamura K, Higashi K: Zinc suppresses apoptosis of U937 cells induced by hydrogen peroxide through an increase of the Bcl-2/Bax ratio. Biochem. Biophys. Res. Commun. 246(2): 364-369, 1998 Fukuchi K, Tomoyasu S, Tsuruoka N, Gomi K: Iron deprivation induced apoptosis in HL60 cells. Febs Lett. 350(1): 139-142, 1994 Furukawa T, Naitoh Y, Kohno H, Tokunaga R, Taketanis S: Iron deprivation decreases ribonucleotide reductase activity and DNA synthesis. Life Sci. 50(26): 2059- 2065, 1992 Gambutti V: Current therapy for thalassaemia in Italy. Ann NY Acad Sci 612: 268-274, 1990 Gadliardini V, Femandez-A P, Lee RKK, Drexler HCA, Rotello RJ, Fishman MC, Yuan J: Prevention of vertebrate neuronal death by the CrmA gene. Science. 263: 826- 828, 1994 Galey JB: Potential use of iron chelators against oxidative damage. Adv Pharmacol 38: 167-203, 1997 251 Galbraith GM, Goust JM, Mercurio SM, Galbraith RM: Transferrin binding by mitogen activated human peripheral blood lymphocytes. Clin. Immunol. Immunopathol. 10(4): 387-395, 1980 Ganeshaguru. K, Hoffbrand.AV, Grady .RW, Cerami A: Effect of various iron chelating agents on DNA synthesis in human cells: Biochemical Pharmacology, 29: 1275-1279, 1980 Giannakis C, Forbes IJ, Zaleswski PD: Ca^^Mg^^-dependent nuclease; tissue distribution, relationship to inter-nucleosomal DNA fragmentation and inhibition by zinc. Biochem. Biophys. Res. Commun. 181: 915-920, 1991 Giardina PJV, Ehlers KH, Engle MA: The effect of subcutaneous desferrioxamine on the cardiac profile of thalassaemia major: A five-year study. Ann NY Acad Sci 445: 282-292, 1985 Giardina PJV, Grady RW, Ehlers KH; Current therapy of Cooleys anaemia. A decade of experience with subcutaneous desferrioxamine. Ann NY Acad Sci 612: 275-285, 1990 Gillis S, Watson J: Biochemical and biological characterization of lymphocyte regulatory molecules V. Identification of an interleukin-2 producing human leukaemia T cell line. J. Exp. Med: 1709-1719, 1980 Goldstein P, Qjcius DM, Young JD: Cell death mechanisms and the immune system. Immunol Rev. 121: 29-65, 1991 Golstein P, Marguet D, Depraetere V: Homology between reaper and the cell-death domains of Fas and TNFR-1. Cell. 81: 185-186, 1995 Gordeuk V, Mukiibi J, Med M: Iron overload in Africa: Interaction between a gene and dietary iron content N Engl J Med 326: 95-100, 1992 Guenal L, Sidoti-de-Fraisse C, Gaumer S, Mignotte B: Bcl-2 and Hsp-27 act at different levels to suppress programmed cell death. Oncogene 15: 347-360, 1997 Gunshin H, Mackenzie B, Berger UV: Cloning and characterisation of a mammalian proton-coupled metal-ion transporter. Nature 388: 482-488, 1997 252 Gutteridge JMC, Rowley DA, Halliwell B: Superoxide dependent formation of hydroxyl radicals and lipid peroxidation in the presence of iron salts. Biochem J 206: 605-609, 1982 Hahn PF, Bale WF, Ross JF: Radioactive iron absorption by the gastointestinal tract- influence of anaemia, anoxia and antecedent feeding. Distribution in growing dogs. J Exp Med 79: 169-188, 1943 Hakem R, Hakem A, Duncan OS, Henderson JT, Woo M, Soengas MS, Elia A, de-la Pompa JL, Kagi D: Differential requirement for caspase-9 in apoptotic pathways in vivo. Cell. 94: 339-352, 1998 Hale AJ, Smith CA, Sutherland LC, Stoneman VEA, Longthome VL, Culhane AC, Williams GT: Apoptosis: Molecular regulation of cell death. Eur. J. Biochem. 236: 1- 26, 1996 Halliwell B, Gutteridge JMC: Biologically relevant metal ion-dependent hydroxyl radical generation-An update. FEBS Lett 307: 108-112, 1992 Haq RU, Wereley JP, Chitambar CR: Induction of apoptosis by iron deprivation in human leukaemic CCRF-CEM cells. Exp. Haematol. 23(5): 428-432, 1995 Harrigan MT, Campbell NF, Bourgeois S: Identification of a gene induced by glucocorticoids in murine T cells: a potential G protein coupled receptor. Mol. Endocrinol. 5(9): 1331-1338, 1991 Harrison PM, Arosia P: The ferritins; Molecular properties, iron storage functions and cellular regulation. Biochim. Biophys. Acta. 1275(3): 161-203, 1996 Hartman RS, Conrad ME, Hartman RE: Ferritin containing bodies in human small intestine. Blood 22; 397-405, 1963 Hentze MW, Kuhn LC: Molecular control of vertebrate iron metabolism: mRNA based regulatory circuits operated by iron, nitric oxide and oxidative stress. Proc Natl Acad Sci USA. 93; 8175-8182, 1996 Hershko C: A study of the chelating agent diethylenetriaminepentacetic acid using selective radioiron probes of reticuloendothelial and parenchymal iron stores. J. Lab and Clin Med 85: 913-921, 1975 253 Hershko C: Determinants of fecal and urinary iron excretion in rats. Blood 51: 415- 423, 1978 Hershko C, Rachmilewitz EA: Mechanisms of desferrioxamine-induced iron excretion in thalassaemia. Brit J Haematol 42: 125-132, 1979 Hershko C, Link G, Pinson A: Modification of iron uptake and lipid peroxidation by hypoxia, ascorbic acid and alpha tocopherol. J. Clin.Lab and Clin Med 110: 355-361, 1987 Hershko C, Weatherall DJ: Iron chelating therapy. CRC Grit Rev Clin Lab Sci 26: 303- 345, 1988 Hershko C; Development of oral iron chelator, LL Lancet 341(8852): 1088-1089, 1993 Hershko C: Control of disease by selective iron depletion , a novel therapeutic strategy utilizing iron chelators. Baillieres Clin Haematol. 7(4): 965-1000, 1994 HiderRC, Kontoghiorghes G, Silver J: Pharmaceutical Compositions, UK patent GB 2118176A, 1982 Hider RC, Ejim L, Taylor PD, Gale R, Huehns E, Porter JB; Facilitated uptake of zinc into human erythrocytes. Biochem Pharmacol.39:1005-1012, 1990 Hider RC, Singh S; Iron chelating agents with clinical potential. Procs Roy. Soc Ed 99B: 137-168, 1992 HiderRC, Choudhury R, Rai BL, Dehkordi LS, Singh S: Design of orally active iron chelators. Acta Haematol 95: 6-12, 1996 Hileti D, Panayiotidis P, Hoffbrand AV: Iron chelators induce apoptosis in proliferating cells. Br. J. Haematol. 89:181-187, 1995 Hirsch S; Gordon S: Polymorphic expression of a neutrophil differentiation antigen revealed by a monoclonal antibody 7/4. Immunogenetics. 18: 229-239, 1983 Hoffbrand AV, Ganeshaguru K, Hooton JWL, Tattersall MHN: Effect of iron deficiency and desferrioxamine on DNA synthesis in human cells. Brit J. Haematol. 33: 517-526, 1976 254 Hoffbrand AV, Bartlett AN, Veys PA, O’Connor NTJ, Kontoghiorghes GJ: Agranulocytosis and thrombocytopenia in patient with Blackfan-Diamond anaemia during oral chelato trial. Lancet 2: 457, 1989 Hoyes KP, Hider RC, Porter JB: Cell cycle synchronisation and growth inhibition by 3-hydroxypyridin-4-one iron chelators in leukaemia cell lines. Cancer Research. 52: 4591-4599, 1992 Hoyes KP, Porter JB: Subcellular distribution of desferrioxamine and hydroxypyiidin- 4-one chelators in K562 cells afffects chelation of intracellular iron pools. BR. J. Haematol 85(2): 393-400, 1993 Hsu H, Xiong J, Goeddel DV: The TNF receptor associated protein TRADD signals cell death and NF kB activation. Cell 81: 495-504, 1995 Hughes TA, Cook PR: Mimosine arrests the cell cycle after cells enter S phase. Exp Cell Res. 222(2): 275-280, 1996 Humecu L, Bardis F, Maroc C, Alavio T, Galindo R, Mamoni P, Chabannon C: Phenotypic, molecular and funcional characterisation of human peripheral blood CD34+/Thy-1 cells. Blood 8793):949-955, 1996 Inaba K, Inaba M, Kinashi T, Tashiro K, Witmer-Pack M, Crowley M, Kaplan G, Valinsky, Romain N: Macrophages phagocytose thymic lymphocytes with productively rearranged T cell receptor alpha and beta genes. J. exp. Med. 168(6): 2279-2294, 1988 Irmler M, Hofmann K, Vaux D, Tschopp J: Direct physical interaction between the Caenorhabditis elegans 'death proteins' Ced-3 and Ced-4. FEBS Lett 406:189-190, 1997 Ishii T, Nakamura M, Yagi H, Soga H, Kayaba S, Gotoh T, Satomi S, Itoh T; Glucocorticoid-induced thymocyte death in the murine thymus; the effect at later stages. Arch Histol Cytol. 60(1): 65-78, 1997 Itoh N, Yonehara S, Ishii A, Yonehara M, Mizushima -I S, Sameshima M, Itase A, Yoshiyuki S, Nagata S; The polypeptide encoded by the cDNA of human cell antigen Fas can modulate apoptosis. Cell. 66: 233-243, 1991 255 Itoh N, Nagata S; A novel protein domain required for apoptosis, mutational analysis of the human Fas antigen. J. Biol. Chem. 268915): 10932-10937, 1993 Iwai K, Klausner RD, Rouault TA: Requirements for iron-regulated degradation of the RNA binding protein, iron regulatory protein 2. EMBO J. 14(21): 5350-5357, 1995 Jacobson MD, Weil M, Raff MC: Role of Ced-3/ICEfamily proteases in staurosporine- induced programmed cell death. J. Cell Biol, 133: 1041-1051, 1996 Jazwinska EC, F*yper WR, Burt MJ: Haplotype analysis in Australian hemochromatosis patients: Evidence for a predominant ancestral haplotype exclusively associated with hemochromatosis. Am J Hum Genet 14: 428-433, 1995 Jiang S, Chow SC, McCabe MJ, Orrenius S: Lack of Ca2+ involvement in thymocyte apoptosis induced by chelation of intracellular Zn 2+. Lab Invest. 73(1): 111-117, 1995 Juckett MB, Shadley JD, Zheng Y, Klein JP: Desferrioxamine enhances the effects of gamma radiation on clonogenic survival and the formation of chromosomal aberrations in endothelial cells. Radiat-Res. 149(4): 330-337, 1998 Karlsson M, Sahlin M, Sjoberg BM: Esherichia coli ribonucleotide reductase; radical susceptibility to hydroxyurea is dependent on the regulatory state of the enzyme. J. Biol Chem 267: 12622-12626, 1992 Kaufman SH, Desnoyers S, Ottaviano Y, Davidson NE, Poirier GO: Specific proteolytic cleavage of poly (ADP) ribose polymerase: an early marker of chemotherapy induced apoptosis. Cancer Res. 53: 3976-3985, 1993 Kayyali R: Analysis of metalloenzyme inhibition by iron chelators developed for the treatment of thalassaemia. PhD Thesis, University of London, 1998 Keberle H: The biochemistry of Desferrioxamine and its relation to iron metabolism. Annals of the New York Acad Sci 119; 758-768, 1964 Kerr JFR, Wyllie AH, Currie AR: Apoptosis: A basic biological phenomenon with wide ranging implications in tissue kinetics. Br. J Cancer 26: 239-257, 1972 256 King LB, Ashwell JD: Signalling for death of lymphoid cells. Curr. Opin. Immunol. 593): 368-373, 1993 King KL, Cidlowski JA: Cell cycle and apoptosis: Common pathways to life and death. J. Cell. Biochem. 58: 175-180, 1995 Kirkhus B, Clausen OP: Cell kinetics in mouse epidermis studied by bivariate DNA/Bromodeoxyuridine and DNA/keratin flow cytometry. Cytometry. 11(2): 253- 260, 1990 Kisielow P, von Boehmer H: Negative and positive seletion of immature thymocytes: timing and the role of the ligand for alpha beta receptor. Semin Immunol 2(1): 35-44, 1990 Klausner RD, Rouault TA, Hartford JT: Regulating the fate of mRNA: The control of cellular iron metabolism. Cell 72: 19-28, 1993 Kontoghiorghes GJ, Aldouri MA, Sheppard L, Hoffbrand AV: l,2-Dimethyl-3 hydroxypyrid-4-one, an orally active chelator for treatment of iron overload. Lancet 1: 1294-1295, 1987 Kontoghiorghes GT: Comparative efficacy and toxicity of desferrioxamine, deferriprone, and other iron and aluminium chelating drugs. Toxicol. Lett. 80(1-3): 1- 18,1995 Kovar J, Seligman P, Gelfund EW: Inhibition of growth of the lymphocyte lines by desferrioxamine under various iron supply conditions. Folia. Biol. Praha 39(1): 29-39, 1993 Kovar J, Stunz LL, Stewart BC, Kiiegerbeckova K, Ashman RF, Kemp JD: Direct evidence that iron deprivation induces apoptosis in murine lymphoma 38C13. Pathobiology. 65(2): 61-68, 1997 Krause DS, Fackler MJ, Civin Cl, Stratford-May W: CD34: Structure, biology and clinical utility. Blood.87:l-13, 1996 Kroemer G, Petit PX, Zamzami N, Vayssiere JL, Mignotte B: The biochemistry of programmed cell death. FASEB j. 9: 1277-1287, 1995 257 Kruse-Jarres JD. J. Trace Hem Hectrolytes Health Dis. 3:1, 1989 Koopman G, Reutelingsperger CPM, Kuijten GAM, Keehnen RMJ, Pals ST, van Oers MHJ: Annexin v for flow cytometry detection of phosphatidylserine expression on B cells undergoing apoptosis. Blood 84: 1415-1420, 1994 Koren G, Kochavi-Atiya Y, Bentur Y: The effects of subcutaneous desferrioxamine administration on renal function in thalassaemia major. Int J Hematol 54: 371-375, 1991 Kretchmar Nguyen SA, Craig A, Raymond KN: Transferrin: The role of conformational changes in iron removal by chelators. J. Am. Chem. Soc 115: 6758- 6764, 1993 Kuida K, Lippke JA, Ku G, Harding MW, Livingston DJ, Su MS, Flavell RA: Altered cytokine export and apoptosis in mice deficient in interleukin 1 p converting enzyme. Science. 267: 2000-2002, 1995 Kuida K, Zheng TS, Na S, Kuan C, Yang D, Karasuyama H, Rakie P, Flavell RA: Decreased apoptosis in the brain and prmature lethality in CPP32-deficient mice. Nature. 384: 368-372, 1996 Kulp KS, Vulliet PR: Mimosine blocks cell cycle progression by chelating iron in asynchronous human breast cancer cells. Toxicol. Appl Pharmacol. 139(2): 356-364, 1996 Kumar S: ICE-like proteases in apoptosis. TIBS. 20:198-202, 1995 Kuo IC, Seitz B, Labree L, McDonnell PJ: Can zinc prevent apoptosis of anterior keratocytes after superficial keratectomy?. Cornea. 16(5): 550-555, 1997 Kurioka S: Synthesis of p-nitophenylphosphorylcholine and the hydrolysis with phospholipase C. J. Biochem. Tokyo. 63(5):678-680, 1968 Kyewsky BA, Schirrmacher A, Allison JP; Antibodies against the T cell receptor/CD3 complex interfere with distinct intrathymic cell-cell interactions in vivo; correlation with arrest of T cell differentiation. Eur. J. Immunol. 19; 857, 1989 258 Kyriakou D, Eliopoulos AG, Papadakis A, Alexandrakis M, Eliopoulos GD: Decreased expression of c-myc oncoprotein by peripheral blood mononuclear cells in thalassaemia patients receiving desferrioxamine. Eur. J. Haematol. 60(1): 21-27, 1998 Lane DP: p53 Guardian of the genome. Nature, 358; 15-16, 1992 Lassman G, Curtis J, Brunhild L, Mason RP, Eling TE: ESR studies on reactivity of protein-derived tyrosyl radicals formed by prostaglandin H synthase and RR. Arch Biochem Biophys. 300(1): 132-136, 1992 Lazebnick YA, Kaufmann SH, Desnoyers S, Poirier GG, Eamshaw WC: Cleavage of poly(ADP-ribose) polymerase by a proteinase with properties like ICE. Nature 371: 346-347, 1994 Lazebnick YA, Takahashi A, Moir RD, Goldman RD, Poirier GG, Kaufmann SH, Eamshaw WC: Studies of the lamin proteinase reveal multiple parallel biochemical pathways during apoptotic execution. Proc Natl Acad Sci. 92: 9042-9046, 1995 Leary AG, Ikebuchi K, Hirai Y, Wong GG, Yang YC, Clark SC, Ogawa M: Synergism between interleukin-6 and interleukin-3 in supporting proliferation of human haematopoietic stem cells; Comparison with interleukin-1 alpha. Blood 71(6): 1759- 1763, 1988 Lee KS, Frank S, Vanderklish P, Arai A, Lynch G: Inhibition of proteolysis protects hippocampal neurons from ishemia. PrOc. Natl. Acad. Sci. USA. 88(16): 7233-7237, 1991 Lee MR, Liou ML, Yang YF, Lai MZ: cAMP analogs prevent activation induced apoptosis of T cell hybridomas. J. Immunol 151(10): 5208-5217, 1993 Lesage S, Steff AM, Philippoussis F, Page M, Trop S, Mateo V, Hugo P: CD4+CD8+ thymocytes are preferentially induced to die following CD45 cross-linking through a novel apoptotic pathway. J. Immunol. 159(10): 4762-4771, 1997 Lee P, Mohammed N, Marshall L, Abeysinghe RD, Hider RC, Porter JB, Singh S: Intraveneous infusion pharmakinetics of desferrioxamine in thalassaemic patients. D. Met. Dispos. 21:640-644, 1993 259 Lederman HM, Cohen A, Lee JWW, Freedman MH, Gelfand EW: Desferrioxamine: a t reversible S-phase inhibitor of human lymphocyte proliferation. Blood 64(3): 748-753, 1984 Leist M, Single B, Kunstle G, Volbracht C, Hentze H, Nicotera P: Apoptosis in the absence of poly (ADP-ribose) polymerase. Biochem. Biophys. Res. Commun. 233(2): 518-522, 1997 Lemoli RM, Tafuri A, Fortuna A, Petrucci MT, Riccardi MR, Catani L, Rondelli D, Fogli M, Leopardi G, Ariola C, Tura S: Cycling status of CD34+ cells mobilised into peripheral blood of healthy donors by recombinant human granulocyte colony stimulating factor. Blood. 89(4): 1189-1196, 1997 Li P, Allen H, Baneijee S, Franklin S, Herzog L, Johnson C, McDowell J, Paskind M, Rodman L, Saif eld J, Towne E, Tracey D, Wardwell S, Wei F-Y, Wong W, Kamen R, Seshadri T: Mice deficient in XL-lb converting enzyme are defective in I production of of mature IL-ip and resistant to endotoxic shock. Cell. 80; 401-411, 1995 Licht S, Gerfen GJ, Stubbe J: Thiyl radicals in ribonucleotide reductases. Science. 271: 477-481, 1996 I Lincz CF: Deciphering the apoptotic pathway: all roads lead to death. Immunol. Cell. Biol. 76(1): 1-19, 1998 Link G, Pinson A, Hershko C: Iron loading of cultured cardiac myocytes modifies sarcolemmal structure and increases lysosomal fragility. J Lab and Clin Med 121; 127- 134,1993 Liu ZG, Smith SW, McLaughlin KA, Schwartz LM, Osborne BA: Apoptotic signals delivered through the T-cell receptor of a T cell hybrid require the immediate early gene Nur77. Nature.367: 281-284, 1994 Longthome VL, Williams GT: Caspase activity is required for commitment to Fas- mediated apoptosis. EMBO J. 16:3805-3812, 1997 Lotem J, Sachs L: Induction of dependence on haematopoietic proteins for viability and receptor upregulation in differentiating myeloid lekaemic cells. Blood 74(2): 579-785, 1989 260 Lowe SW, Schmitt EM, Smith SW, Osborne BA, Jacks T: p53 is required for radiation induced apoptosis in mouse thymocytes. Nature. 362: 847-849, 1993 Lowenberg B, Touw IP: Hematopoietic growth factors and their receptors in acute leukaemia. Blood 81(2): 281-292, 1993 Lucas JJ, Szepesi A, Domenico J, Takase K, Tordai A, Terada N, Gelfand EW: Effects of iron depletion on cell cycle progression in normal human T lymphocytes: selective inhibition of the appearance of the cychn A associated component of the p33cdk2 kinase. Blood. 86(6): 2268-2280, 1995. Maclean KH, Lynagh G, Porter JB: Mechanisms of induction of thymocyte apoptosis by iron chelators. European Iron club meeting, Hamburg, Germany, 1995 Maclean KH, Porter JB: Contrasting apoptotic induction in bone marrow precursors and thymocytes by hydroxyurea and iron chelators. American Socity of Haematology meeting, 1996 Maclean KH, Porter JB: Molecular mechanisms of chelator-induced thymocyte apoptosis. International symposium of iron in biology and medicine meeting, 1997 Macdonald HR, Lees RK: Programmed death of autoreactive thymocytes. Nature 343(6259): 642-644, 1990 Macfarlane M, CainK, Sun XM, Alnemri ES, Cohen GM: Processing/activation of at least 4 interleukin 1 beta converting enzyme-like protease occurs during the execution phase of apoptosis in human monocytic tumour lines. J. Biol. Chem 137(2): 469-479, 1997 Marachi T, Takano E, Maki M, Adachi Y, Hatanaka M: Cloning and expression of the genes for calpains and calpastatins. Biochem. Soc. Symp. 55: 29-44, 1989 Marini M, Musiani D: Micromolar zinc affects endonucleolytic activity in hydrogen peroxide-mediated apoptosis. Exp. Cell. Res. 239(2): 393-398, 1998 Martin SJ, Cotter TG: Ultraviolet B irradiation of human leukaemia HL60 cells in vitro induces apoptosis. Int J Radiat Biol. 59: 1001, 1991 261 Martin SJ, Mazdai G, Strain JJ, Cotter TG, Hannigan BM; Programmed cell death (apoptosis) in lymphoid and myeloid cell lines during zinc deficiency. Clin Exp Immunol. 83; 338-343, 1991 Martin SJ, Lennon SV, Bonham AM, Cotter TG: Induction of apoptosis (programmed cell death) in human leukaemia HL60 cells by inhibition of RNA or protein synthesis. J. immunol. 145: 1859-1867, 1990 Martin SC, Reutelingsperger A, McGon J, Rader R, van Schie D, LaFace, Green D: Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus: inhibition by overexpression of Bcl2 and Abl. J. Exp. Med. 182: 1545-1556, 1995 Martin SJ, Bonham AM, Cotter TG: The involvement of RNA and protein synthesis in programmed cell death (apoptosis) in human leukaemic HL60 cells. Biochem-Soc- Trans. 18(4): 634-636, 1990 Martin SJ, O'Brien GA, Nishioka WK, McGahon AJ, Mahboubi A, Saido TC, Green DR: Proteolysis of fodrin (non-erythroid spectrin) during apoptosis. J. Biol. Chem. 270: 6425-6428, 1995 Mathieu J, Ferlat S, Ballester B, Platel S, Herodin F, Chancerelle Y, Mestries JC. Kergonou JF: Radiation-induced apoptosis in thymocytes; inhibition by diethyldithiocarbamate and zinc. Radiat. Res. 146(6): 652-659, 1996 Matsubara K, Kubota M, Adachi S, Kuwakado K, Hirota H, Wakazono Y, Akiyama Y, Mikawa H: Different mode of cell death induced by calcium ionophore in human leukaemia cell lines; possible role of constituative endonuclease. Exp. Cell Res. 210: 19-25, 1994 Matsui D, Klein J, Hermann C: Relationship between the pharmacokinetics and iron excretion pharmacodynamics of the new oral iron chelator 1,2 dimethyl-3 hydroxypiridin-4-one in patients with thalassaemia. Clin Pharmacol Ther 50: 294-298, 1991 Mayani H, Lansdorp PM: Thy-1 expression is linked to funcional properties of primative hematopoietic progenitor cells from human umbilical cord blood. Blood 83(9): 2410-2417, 1994 262 McCabe MJ, Jiang S, Orrenius S: Chelation of intracellular zinc triggers apoptosis in mature thymocytes. Lab Invest. 69: 101-110, 1993 McCarthy NJ, Whyte MK, Gilbert CS, Evan GI: Inhibition of ced-3/ICE-related proteases does not prevent cell death induced by oncogenes, DNA damage or the Bcl-2 homologue, Bak. J.Cell. Biol. 136(1): 215-227, 1997 McCord JM: Free radicals and inflammation: Protection of synovial fluid by suoeroxide dismutase. Science 185: 529-531, 1974 McCord JM: Iron, Free Radicals and oxidative injury. Semin-Haematol. 35(1): 5-12, 1998 Mehta J, Singhal S, Revankar R, Walvankar A, Chablani A: Fatal systemic lupus erythematous in patients taking oral iron chelator LI. Lancet. 337(8736): 298, 1991 Mehta J, Singhal S, Mehta BC: Oral iron chelator LI and autoimmunity. Blood 81: 1971-1972, 1993 Meneghini G: Iron homeostasis, oxidative stress and DNA damage. Free-Radic-Biol- Med. 23(5): 783-792, 1997 Meisenholder GW, Martin SJ, Green DR, Nordberg J, Babior BM, Gottlieb: Events in apoptosis. Acidifiaction is downstream of protease activation and BcI-2 protection. J.Biol Chem. 271: 16260-16263, 1996 Metcalf D: In vitro cloning techniques for haemopoietic cells: Clinical applications. Ann. Intern. Med 87(4); 483-488, 1977 Metcalf D: Haemopoietic growth factors Med. J. Aust 148(10): 516-519, 1988 Metcalf D: The molecular control of division , differentiation conunitment and maturation in haemopoietic cells. Nature 339: 27-30, 1989 Metcalf D and Nicola NA: The haemopoietic colony-stimulating factors. Cambridge University Press, 1995 Migliorati G, Nicoletti I, Pagliacci MC, D ’AdamioL, Riccardi C: Interleukin-2 induces apoptosis in mouse thymocytes. Cell Immunol. 146(1): 52-61, 1993 263 Mignotte B, Vayssiere JL: Mitochondria and apoptosis. Eur J Biochem 252: 1-15, 1998 Mills K, Armitage R, Worman C: An indirect rosette technique for the identification and seperation of human lymphocyte populations by monoclonal antibodies: A comparison with immuno-fluorescenve techniques. Immunol Lett. 6 : 241-246, 1983 Miura M, Zhu H, Rotello R, Hartweig EA, Yuan J: Induction of apoptosis in fibroblasts by 11^ ip converting enzyme, a mammalian homolog of the C.elegans cell death gene Ced-3. Cell 75: 653-660, 1993 Miura M, Yuan J: Regulation of programmed cell death by interleukin 1 beta converting enzyme family of proteases. Adv. Exp. Med. Biol. 389: 165-172, 1996 Modell CB, Beck J: Long-term desferrioxamine therapy in thalassaemia. Ann NY Acad Sci 232: 201-210, 1974 Modell B: Advances in the use of iron-chelating agents for the treatment of iron overload. Prog Haematol. 11: 267, 1979 Morgan EH: Inhibition of reticulocyte iron uptake by NH 4 CI and CH 3 NH2 . Biochim Biophys Acta 642:119-134, 1981 Mousa SA, Ritger RC, Smith RD: Efficacy and safety of desferrioxamine conjugated to hydroxyethyl starch. J. Cardiovascu. Pharmacol. 19(3): 425-429, 1992 Murphy KM, Heimbeger AB, Loh DY: Induction by anigen of intrathymic apoptosis of CD4-i-CD8-i-TCRLo thymocytes in vivo. Science 250(4988): 1720-1723, 1990 Nagata S: Apoptosis by death factor. Cell 88(3); 355-365, 1997 Nagata S, Golstein P: The Fas death factor. Science.267; 1449-1456, 1995 Neamati N, Fernandez A, Wright S, Kiefer J, McConkey DJ; Degradation of lamin B1 precedes oli gonucleosomal DNA fragmentation in apoptotic thymocytes and isolated thymocyte nuclei. J. Immunol. 154(8): 3788-3795, 1995 Neldner KH, Hambidge KM, Walravens PA. Int J Dermatol. 17; 380, 1978 264 Nicoletti I, Migliorati G, Pagliacci MC, Grignani F, Riccardi C: A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry. J. Immunol. Meth. 139: 271-279, 1991 Nicholson DW, Ali A, Thomberry NA, Vaillancourt JP, Ding CK, Gallant M, Gareau Y, Griff en PR, Labelle M, Lazebnick YA, Munday NA, Raju SM, Smulson ME, YaminTT, Yu VL, Miller DK: Identification and inhibition of the lCE/Ced-3 protease necessary for mammalian apoptosis. Nature376: 37-43, 1995 Ning ZQ, Norton JD, Li J, Murphy JT: Distinct mechanisms for rescue from apoptosis in ramos human B cells by signalling through CD40 and interleukin-4 receptor; role for inhibition early response gene, Berg36. Biochem. Soc. Trans. 25(2): 306s, 1997 Nyholm S, Mann GJ, Johansson AG, Bergeron RJ, Graslund A, Thelander L: Role of ribonucleotide reductase in inhibition of mammalian cell growth by potent iron chelators. J. Biol. Chem 268(35): 26200-26205, 1993 Nyholm S, Thelander L, Grasluund A: Reduction and loss of iron centre in the reaction of the small subunit of mouse ribonucleotide reducatase with hydroxyurea. Biochemistry .32: 11569-11574, 1995 Oberhammer F, Fritsch G, Parelka M, Froschl G, Tiefenbacher R, Purchio T, Schulte- Hermann R: Induction of apoptosis in cultured hepatocytes and in the regressing liver by transforming growth factor beta 1 occurs without activation of an endonuclease. Toxicol Lett. 64-65, 1992 Oberhammer F, Wilson JW, Dive C, Morris ID, Hickman JA, Wakehng AE, Walker PR, Sikorska M: Apoptotic death in epithelial cells: Cleavage of DNA to 300 and/or 50kb fragments prior to or in the absence of intemucleosomal fragmentation. EMBO J. 12: 3679-3684, 1993 Ochiai E, Mann GJ, Graslund A, Thelander L: Tyrosyl free radical formation in the small subunit of mouse ribonucleotide reductase. J. Biol. Chem 265(26): 15758- 15761, 1990 Olivieri NF, Koren G, Hermann C: Comparison of oral iron chelator LI and desferrioxamine in iron loaded patients. Lancet 336: 1275-1279, 1990 265 Olivieri NF, Nathan DG, MacMillan JH: Survival in medically treated patients with homozygous p-thalassaemia. N Engl J Med 331: 574-578, 1994 Olivieri NF, Brittenham GM, Matsui D, Berkovitch M, Blendis LM, Cameron RG, McLelland RA, Lui PP, Templeton DM, Koren G: Iron chelation therapy with oral deferiprone in patients with thalassaemia major. New Engl J Med 332: 918-922, 1995 OllagnierS, MulliezE, Schmidt PP, Eliasson R, Gaillard J, Deronzier C, Bergman T, Graslund A, Reichard P, Fontecave M: The anaerobic Esherichia-Coli ribonucleotide reductase: subunit structure and iron sulfur center. J.Biol. Chem 271(16): 9410-9416, 1996 Orrenius S, McCabe MJ jr, Nicotera P: Ca2+-dependent mechanisms of cytotoxicity and programmed cell death. Toxicol. Lett. 64-65: spec no. 357-364, 1992 Ottolenghi AC: Phospholipase C from bacillus cereus, a zinc requiring metalloenzyme. Biochim-Biophys-Acta. 106(3): 510-518, 1965 Paramanantham R, Sit KH, Bay BH: Adding Zn2-f induces DNA fragmentation and cell condensation in cultured human chang liver cells. Biol Trace Elem Res.58: 135- 147, 1997 Pardee AB: G l events and regulation of cell proliferation. Science 246(4930): 603-608, 1989 Park DS, Stefanis L, Yan CYl, Farinelli SE, Greene LA: Ordering the cell death pathway. J.Biol Chem. 271: 21898-21905, 1996 Parkes JG, Hussain RA, Ohvieri NF, Templeton DM: Effects of iron loading on uptake, spéciation and chelation of iron in cultured myocardial cells. J. Lab. Clin. Med 122(1): 36-47, 1993 Pattanapanyasat K, Webster HK, Tongtawe P, Kongharoen P, Hider RC: Effect of orally active hydroxypiridinone iron chelators on human lymphocyte function. Br. J. Haematol. 82(1): 13-9, 1992 Peitsch MC, Muller C, Tschopp J: DNA fragmentation during apoptosis is caused by frequent single-strand cuts. Nucleic Acids Res. 21: 4206-4209, 1993 266 Perandones CE, Illera VA, Peckham D, Stunz LL, Ashman RF: Regulation of apoptosis in vitro in mature murine spleen T cells. J. Immunol. 151(7): 3521-3529, 1993 Perry DK, Smyth MJ, Stennicke HR, Salvesen GS, Duriez P, Poirier GG, Hannun YA: Zinc is a potent inhibitor of the apoptotic protease, caspase-3.30; 18530-18533, 1997 Pette M, Gold R, Pette DF, Hartung HP, Toyka KV: Mafostamide induces DNA fragmentation and apoptosis in human T lymphocytes; A possible mechanism of its immunosuppressive action. Immunopharmacology. 30(1): 59-69, 1995 Philpott NJ, Prue RL, Marsh JC, Gordon-Smith EC, Gibson FM: G-CSF-mobilised CD34 peripheral blood stem are significantly less apoptotic than unstimulated peripheral blood CD34 cells: Role of G-CSF as a survival factor. Br. J. Haematol. 97(1): 146-152, 1997 Pippard MJ, Johnson DK, Finch CA: A rapid assay for evaluation of iron chelating agents in rats. Blood 58(4): 685-691, 1981 Pippard MJ, Johnson DK, Finch CA: Hepatocyte iron kinetics in the rat explored with an iron chelator. Br. J. Haematol. 52(2): 211-224, 1982 Pippard MJ, Jackson MJ, Hoffman K, Petrou M, Modell CB: Iron chelation using subcutaneous infusions of diethyl tiiaminepenta-acetic acid (DTPA). Scand. J. Haematol, 36: 466-472, 1986 Pippard MJ: Desferrioxamine induced iron excretion in humans. Ballieres Clin Haematol 2: 323, 1989 Pippard MJ: Secondary iron overload. In Iron metabolism in health and disease. Brock JH, Halliday JW, Pippard MJ, Powell LW (Eds), WB Saunders Co. Ltd. London: 272-309, 1994 Ponka P, Beaumont C, Richardson DR: Function and regulation of transferrin and ferritin. Seminars in haematology, vol 35 no.l(January): 35-54, 1998 Ponka P, Richardson DR: Can ferritin provide iron for haemoglobin synthesis?. Blood 89; 2611-2612, 1997 267 Porter JB, Gyparki M, Burke LC, Heuhns ER, Sarpong P, Saez V, Hider RC: Iron mobilisation from hepatocyte monolayer cultures by chelator the importance of membrane permeability and the iron binding constant. Blood 72: 1497-1503, 1988 Porter JB, Huehns ER, HiderRC: The development of iron chelating drugs. Baillieres clinical haematology 2: 257-292, 1989 Porter JB, Hoyes KP, Abeysinghe RD, Brookes PN, Huehns ER, Hider RC: Comparison of the subacute toxicity and efficacy of 3-hydroxypyridin-4-one iron chelators in overload and nonoverload mice. Blood 78 (10): 2727-2734, 1991 Porter JB, Abeysinghe RD, Hoyes KP, Barra C, Huehns ER, Brooks PN, Blackwell MP, Araita M, Brittenham G, Singhe S et al. Contrasting interspecies efficacy and toxicology of 1, 2 diethyl-3-hydroxypiridin-4-one, CP94, relates to differing metabolism of the iron binding site. Br. J. Haematol. 85(1): 159-168, 1993 Porter JB, Lynagh GR, Hider RC: Iron chelators promote apoptosis in thymocytes and proliferating leukaemic cells. Br. J Haematol 87 (suppl 1): 114, 1994 Porter JB: Development of primary hepatocyte cultures for the preclinical evaluation of iron chelators. PhD Thesis, University of Cambridge, 1998 Prochownik EV, Kukowska J, Rodgers C: cMyc anti sense transcripts accelerate differentiation and inhibit G 1 progression in murine erythroleukaemic cells. Mol. Cell. Biol 8(9): 3683-3695, 1988 Pronk GJ, Ramer K, Amiri P, Williams LT: Requirement for an ICE-like protease for the induction of apoptosis and ceramide generation by REAPER.Science. 271:808-810, 1996 Propper RD, Cooper B, Ruto RR, Neinhuis AW, Anderson WF, Bunn HF, Rosenthal A, Nathan DG: Continuous subcutaneous administration of desferrioxamine in patients with iron overload. N.Engl.J.Med. 297(8): 418-423, 1977 Propper RD, Nathan DC: Clinical removal of iron. Ann Rev Med 33: 509-519, 1982 Provinciali M, Di-Stefano G, Fabris N: Dose dependent opposite effect of zinc on apoptosis in mouse thymocytes. Int. J. Immunopharmacol. 17(9): 735-744, 1995 268 Provinciali M, Di-Stefano G, Stronati S: Flow cytometric analysis of CD3/TCR complex, zinc, and glucocorticoid-mediated regulation of apoptosis and cell cycle distribution in thymocytes from old mice. Cytometry. 32(1): 1-8, 1998 Que L Jr.: Oxygen activation at the diiron enter of ribonucleotide reductase. Science 253(5017): 273-274, 1991 Raff MC: Social controls on cell survival and cell death. Nature 356(6368): 397-400, 1992 Ratan RR, Murphy TH, Baraban JM: Oxidative stress induces apoptosis in embyonic corticol neurons. J. Neurochem. 62: 376-379, 1994 Reed JC, Zha H, Aime SC, Takayama S, Wang HG: Structure function analysis of Bcl-2 family proteins, regulators of programmed cell death. Adv Exp Med Biol. 406: 99-112, 1996 Reems JA and Torok-Storb B: Cell cycle and functional differences between CD34+/CD38hi and CD34+/CD381o human marrow cells after in vitro cytokine exposure. Blood 85: 1480-1487, 1995 Reichard P: From RNA to DNA, Why so many ribonucleotide reductases. Science 260(5115): 1773-1777, 1993 Renton FJ, Jeitner TM: Cell-cycel dependent inhibition of the proliferation of human neural tumour cell lines by iron chelators. Biochem Pharmacol. 51(11): 1553-1561, 1996 Rice WG, Hillyer CD, Harten B, Schaeffer CA, Dorminy M, Lackey DA, Kirsten E, Meddeleyev J, Buki KG, Hakam A, Kun E: Induction of endonuclease mediated apoptosis in tumour cells by c-nitroso-substitated ligands of poly(ADP-ribose) polymerase. Proc Natl Acad Sci USA. 89: 7703-7707, 1992 Richardson DR, Ponka P: The molecular mechanisms of the metabolism and transport of iron in normal and neoplastic cells. Biochim Biophys Acta 1331: 1-40, 1997 Risdon RA, Barry M, Flynn DM: Transfusional iron overload: The relationship between tissue iron concentration and hepatic fibrosis in thalassaemia. J. Pathol 116: 83-95, 1975 269 Rothenburg BE, Voland JR: p2 knockout mice develop parenchymal iron overload: A putative role for class 1 genes of the major histocompatibility complex in iron metabolism. Proc Natl Acad Sci USA 93: 1529-1534, 1996 Rouault TA, Klausner RD: Iron-sulfur clusters as biosensors of oxidants and iron. TIBS. 21: 174-177, 1996 Rusten LS, Jacobsen SB, Kaalhus O, Veiby OP, Funderud S, Smeland EB: Functional differences between CD38 and DR subfractions of CD34+ bone marrow cells. Blood 84(5): 1473-1481, 1994 Saido TC, Yokata M, Nagao S, Yamaura I, Tani E, Tsuchiya T, Suzuki K, Kawashima S: Spatial resolution of fodrin proteolysis in post-ischaemic brain. J. Biol. Chem. 268(33): 25239-25243, 1993 Saido TC, Sorimachi H, Suzuki K: Calpain: New perspectives in molecular diversity and physiological pathological involvement. FASEB. 8(11): 814-822, 1994 Sakabe I, Paul S, Dansithong W, Shinozawa T: Induction of apoptosis in Neuro-2A cells by Zn2+ chelating. Cell-Struct. Funct. 23(2): 95-99, 1998 Sato N, Iwata S, Nakamura K, Hori T, Mori K, Yodoi J: Thiol-mediated redox regulation of apoptosis. Possible roles of cellular thiols other than glutathione in T cell apoptosis. J. Immunol. 154: 3194-3203, 1995 Schade AL, Caroline L: An iron-binding component of human blood plasma. Science 104:340-341, 1946 Schievella AR, Chen JH, Graham JR, Lin-L L: MADD, a novel death domain protein that interacts with the type 1 tumour necrosis factor receptor and activates mitogen- activated protein kinase. J.Biol Chem 272; 12069-12075, 1997 Schlegal RA, Callahan M, Krahlings S, Pradhan D, Williamson P: Mechanisms for recognition and phagocytosis of apoptotic lymphocytes by macrophages. Adv Exp Med Biol.406: 21-28, 1996 Scott JA, Khaw BA, Locke E, Haber E, Homey C: The role of free radical mediated processes in oxygen related in cultured myocardial cells. Circ. Res. 56(1): 72-77, 1985 270 Selden C, Owen M, Hopkins JM, Peters TJ: Studies on the concentration and intracellular localisation of iron proteins in liver biopsy specimens from patients with iron overload with special reference to their role in lysosomal disruption. Br. J. Haematol. 44(4): 593-603, 1980 Sellins KS, Cohen JJ: Gene induction by irradiation leads to DNA fragmentation in thymocytes. J. Immunol. 139: 3199-3206, 1987 Seshagiri S, Miller LK: Caenorhabditis elegans Ced-4 stimulates Ced-3 processing and Ced-3 induced apoptosis. Curr Biol. 7: 455-460, 1997 Shaham S, Horvitz H: Caenorhabditis Elegans neurons may contain both cell-death protective and killer acivities. Genes & Dev. 10: 578-591, 1996 Shalev O, Repka T, Goldfarb L; Deferiprone (LI) chelates pathologic iron deposits from membranes of intact thalassaemic and sickle red blood cells both in vitro and in vivo. Blood 86: 2008-2013, 1995 Shankar AH, Prasad AS: Zinc and immune function: the biological basis of altered resistance to infection. Am J Clin Nutr 68: 447s-463s, 1998 Shumaker DK, Vann LR, Goldberg MW, Allen TD, Wilson KL: TPEN, a Zn2+/Fe2+ chelator with low affinity for Ca2-f, inhibits lamin assembly, destabalizes nuclear architecture and may independently protect nuclei from apoptosis in vitro. Cell Calcium. 23(2-3): 151-164, 1998 Silverman RH, Watling D, Balkwill ER, Trowsdale J, Kerr IM: The PPP(A2’P) nA and protein kinase systems in wild-type and interferon-resistant daudi cells. Eur. J. Biochem. 126(2): 333-341, 1982 Simon M, Bourel M, Genetet B: Idiopathic Hemochromatosis: Demonstration of recessive transmission and early detection by family HLA typing. N Engl J Med 297: 1017-1021, 1977 Singh S, Hider RC: Colorimetric detection of the hydroxyl radical: Comparison of the hydroxyl radical generating ability of various iron complexes. Analytical Biochem 171: 47-51, 1988. 271 Skikne B, Baynes RD: Iron absorption in Brock JH, Halliday JW, Pippard MJ (Eds): Iron metabolism in health and disease. London, Saunders, 1994 Slee EA, Zhu H, Chow SC, Macfarlane M, Nicholson DW, Cohen GM; Benzyloxy carbonyl-val-ala-asp (Ome) fluoromethylketone (ZVAD-fmk) inhibits apoptosis by blocking the processing of CPP32. Biochem J.315: 21-24, 1996 Smeland EB, Funderud S, Kralheim G, Gaudemack G, Rasmussen AM, Rusten L, Wang MV, Tindle RW: Isolation and characterisation of human hematopoietic progenitor cells; An effective method for positive selection of CD34+ cells. Leukaemia 698); 845-52, 1992 SmdhA, Morgan WT: Hemopexin-mediated transport of heme into isolated rat hepatocytes. J.Biol. Chem 256: 10902-10909, 1981 Smith CA, Williams GT, Kingston R, Jenkinson EJ, Owen JJT: Antibodies to CD3/T- cell receptor complex induce death by apoptosis in immature T cells in thymic culture. Nature. 337: 181-184, 1989 Smith RS: Chelating agents in the diagnosis and treatment of iron overload in thalassaemia. Ann NY Acad Sci 119: 776-788, 1964 Solis-Recendez MG, Perani A, D'Habit B, Stacey GN, Maugras M: Hybridoma cell cultures continuously undergo apoptosis and reveal a novel lOObp DNA fragment. J. Biotechnol. 38: 117-127, 1995 Sorimachi H, Saido TC, Suzuki K: New era of calpain research. Discovery of tissue- specific calpains. Febs Lett. 343(1); 1-5, 1994 Spector M, Desnoyers S, Hoeppner D, Hengartner M: Interaction between the C. Elegans cell-death regulators Ced-9 and Ced-4. Nature 385: 653-656, 1997 Squier MK, Miller ACK, Malkinson AM, Cohen JJ: Calpain activation in apoptosis. J.Cell. Physiol. 159: 229-237, 1994 Squier MK, Cohen JJ: Calpain, an upstrean regulator of thymocyte apoptosis. J. Immunol. 158(8): 3690-3697, 1997 272 Srinivasula SM, Ahmad M, Ferdandes-Alnemri T, Litwack G, Alnemri ES: Molecular ordering of the Fas-apoptotic pathway. The Fas/APO-1 protease Mch5 is a CrmA inhibitable protease that activates multiple Ced-3/ICB like cysteine proteases. Proc Natl AcadSci.93; 14486-14491, 1996 Srour EF, Zanjani ED, Cometta K, Traycroff CM, Flake AW, Hedrick M, Brandt JE, Leemhuis T, Hoffman R: Resistance of human multi-lineage, self-renewing lymphohaemopoietic stem cells in chimeric sheep. Blood 82(11): 3333-3342, 1993 Stanger BZ, Leder P, Lee-H T, Kim E, Seed B: RIP; a novel protein conbtaining a death domain that interacts with Fas/APO-1 (CD95) in yeast and causes cell death. Cell 81: 513-523, 1995 Stefanelli C, Bonavita F, Stanic I, Famiggia G, Falcien E, Robuffo I, Pignatti C, Muscaii C, Rossoni C, Guamieri C, Caldavera CM: ATP depletion inhibits glucocorticoid induced thymocyte apoptosis Biochem J. 322(pt 3): 909-917, 1997 Steiner P, Philipp A, Lukas J, Godden-Kent G, Pagano M, Mittnacht S, Bartek J, Eilers M: Identification of a myc-dependent step during the formation of active G l cyclin cdk complexes. EMBO. J. 14(19): 4814-4826, 1995 Steller H: Mechanisms and genes of cellular suicide. Science. 267: 1445-1449, 1995 Stumpf JL, Townsend KA: Other anaemias. In. Clinical pharmacy and therapeutics. Herfmdal ET, Gourley DR, Hart LL (Eds), Williams and Wilkins: 191-209, 1992 Sun XM, Dinsdale D, Snowden RT, Cohen GM, Skilleter DN: Characterisation of apoptosis in thymocytes isolated from dexamethasone-treated rats. Biochem Pharmacol. 44(11): 2131-2137, 1992 Sun DY, Jiang S, Zheng LM, Ojcius DM, Youbg JD: Seperate metabolic pathways leading to DNA fragmentation and apoptotic chromatin condensation. J. Exp. Med. 179: 559-567, 1994 Sutherland R, Delia D, Schneider C, Newman R, Kemshead J, Greaves M: Ubiquitous cell surface glycoprotein on tumour cells is proliferation associated receptor for transferrin. Proc Natl Acad Sci USA. 78: 4515, 1981 273 Swat W, Ignatowicz L, Kisielow P: Detection of apoptosis of immature CD4f8+ thymocytes by flow cytometry. J Immunol Meth. 137: 79-87, 1991 TaetleR, Honeysett JM, Trowbridge I: Effects of anti-transferrin recepeptor antibodies on growth of normal and malignant myeloid cells. Int. J. Cancer 32(3): 343-349, 1983 Takahashi S, Onedera K, Motohashi H, Suwabe N, Hayashi N, Yanai N, Nabesima Y, Yamamoto M: Arrest in primitive erythroid cell development caused by promoter- specific disruption of the GATA-1 gene. J. Biol. Chem. 272(19): 12611-12615, 1997 Takano YS, Harmon BV, Kerr JF: Apoptosis induced by mild hypothermia in human and murine cell lines: a study using electron microscopy and DNA gel electrophoresis. J. Pathol. 163(4): 329-336, 1991 Taylor CG, Bettger WJ, Bray TM: Effect of dietary zinc or copper deficiency on the primary free radical defence system in rats. J. Nutrit. 118(5): 613-621, 1988 Telford WG, Frankar PJ; Preferential induction of apoptosis in mouse CD4+CD8+ alpha beta TCRlo CD3 epsilonlo thymocytes by zinc. J. Cell. Physiol. 164(2): 259- 270, 1995 Telford WG, Franker PJ: Zinc reversibly inhibits steroid binding to murine glucocorticoid receptor. Biochem Biophys Res Commun. 238: 86-89, 1997 Templeton DM: Therapeutic use of chelating agents in iron overload. In. Toxicology of metals; biochemical aspects. Goyer RA and Cherian MG (Eds), Springer-Verlag, Berlin: 305-331, 1995 Testa NG, Hendry JH, Molineux G: Long term bone marrow damage in experimental systems and in patients after radiation or chemotherapy. Anti-Cancer. Res. 5(1): 101- 110, 1985 Tewari M, Dixit VM: Fas and tumour necrosis factor induced apoptosis is inhibited by the poxvirus CrmA gene product. J.Biol Chem 270: 3255-3260, 1995 Thelander L, Reichard P: Reduction of ribonucleotides. Ann. Rev. Biochem. 48: 133- 158, 1979 274 Thiel EC: Ferritin: Structure, gene regulation and cellular function in animals, plants and microorganims. Ann Rev Biochem 56: 289-315, 1987 Thiel EC: Iron regulatory elements (IRE's): A family of mRNA non-coding sequences. Biochem J 304: 1-11, 1994 Thompson CB: Apoptosis in the pathogenesis and treatment of disease. Science 267(5203): 1456-1462, 1995 Thomberry NA, Bull HG, Calaycay JR, Chapman KT, Howard AD, Kostura MJ, Miller DK, Molineaux SM, Weidner JR, Aunins J, EllistonKO, Ayala JM, Casano FJ, Chin J, Ding GJ, Egger LA, Gaffney EP, Limjuco G, Palyha OC, Raju SM, Rolando AM, Salley JP, YaminTT, Lee TD, Shively JE, McCross M, Mumford RA, Schimidt JA, Tocci MJ: A novel heterdimeric cysteine protease is required for interleukin 1 p processing in monocytes. Nature. 356: 768-774, 1992 Till JE, McCulloch EA: A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiat. Res. 14: 213, 1961 Tiwari S: Cyclin dependent kinase regulation in haemopoietic cells. PhD Thesis, University of London, 1998 Treves S, Trentini PL, Ascanelli M, Bucci G, Di Virgilio F: Apoptosis is dependent on intracellular zinc and independent on intracellular calcium in lymphocytes. Exp Cell Res.211:339-343, 1994 Trubiani O, Antonucci A, Palka G, Di-Palmio R; Programmed cell death in peripheral myeloid precursor cells in Downs patients; effect of zinc therapy. Ultrastruct. Pathol. 20(5): 457-462, 1996 Tsien RY, PozzanT, RinkTJ: Calcium homeostasis in intact lymphocytes; cytoplasmic free calcium monitored with a new intracellularly trapped fluorescent indicator. J Cell Biol. 94: 325-334, 1982 Turnbull A: Iron absorption. In: Iron in biochemistry and medicine. Jacobs A and Worwood M (Eds), Academic Press, London and New York: 369-395, 1974 275 Umbeit JN, Conrad ME, Moore EG: Paraferritin: A protein complex with ferrireducatase activity is associated with iron absorption in rats. Biochemistry. 35: 6460-6468, 1996 Umbreit JN, Conrad ME, Moore EG, Latour LF: Iron absorption and cellular transport: The mobilferrin/paraferritin paradigm. Seminars in Haematology, vol 35, n o.l (January): 13-26, 1998 Uzel C, Conrad ME: Absorption of heme iron. Semin. Haematol. 35: 27-34, 1998 Vanags DM, Pom-Ares MI, Coppola S, Burgess DH, Orrenius S: Protease involvement in fodrin cleavage and phosphatidylserine exposure in apoptosis. J. Biol. Chem. 271(49): 31075-31085, 1996 Vanderlaan M, Thomas CB: Characterisation of monoclonal antibodies to bromodeoxyuridine. Cytometry. 6(6): 501-505, 1985 Vayssiere JL, Petit PX, Risler Y, Mignotte B: Commitment to apoptosis is associated with changes in mitochondrial biogenesis and activity in cell lines conditionally immortalized with simian virus 40. Proc Natl Acad Sci USA. 91: 11752-11756, 1994 Vincent JB, Olivier-Iilley GL, Averill BA; Proteins containing oxo-bridged dimuclear iron centres; a bioinorganic perspective. Chem. Rev. 90: 1447-1467, 1990 Vincenz C, Dixit VM: Fas-associated death domain protein interleukin 1 p converting enzyme 2 (FLICE-2), an ICE/Ced-3 homologue, is proximally involved in CD95 and p55-mediated death signalling. J. Biol Chem. 272: 6578-6583, 1997 Walker PR, Smith C, YoudaleT, Leblanc J, Whitfield JF, Sikorska M: Topoisomerase II reactive drugs induce apoptosis in thymocytes. Cancer Res 51(40; 1078-1085, 1991 Wang KK, Roufogatin BD, Villalobo A: Calpain 1 activates Ca2+ transport by the reconstituted erthyrocyte Ca2+ pump. J. Memb. Biol 112(3): 233-245, 1989 Wang S, Miuyra M, Jung Y, Zhu H, Li E, Yuan J: Murine caspase-11, an ICE interacting protease, is essential for the activation of ICE. Cell. 92:501-509, 1998 Wang ST, Jin KL, Li TH: The inhibitory effect of desferrioxamine on DNA synthesis in human lymphocytes. Chin. Med. J. Engl. 102(12): 902-905, 1989 n 76 Waring P: DNA fragmentation induced in macrophages by gliotoxin does not require protein synthesis and is preceded by raised inositol triphosphate levels. J. Biol. Chem. 265(24): 14476-14480, 1990 Waring P, Egan M, Braithwaite A, Mullbacher A, Sjaarda A: Apoptosis induced in macrophages and T blasts by the mycotoxin sporidesmin and protection by Zn2+ salts: Int. J. Immunopharmacol. 12(4): 445-457, 1990 Watts MJ, Sullivan AM, Ings SJ, Leverett D, Peniket AJ, Perry AR, Williams CD, Devereux S, Goldstone AH, Linch DC: Evaluation of clinical scale CD34+ cell purification; experience of 71 immunoaffinity column procedures. Bone marrow transplant. 20: 157-162, 1997 Watts MJ: Human haemopoietic progenitor cell mobilization. PhD thesis. University of London, 1999 Weatherall DJ, Clegg JB: The thalasaaemia syndromes, 3rd edition Oxford: Blackwell Scientific, 1981 Weatherall DJ: Disorders of the synthesis or function of haemoglobin. Oxford textbook of medicine Vol. 3. Weatherall DJ, Ledingham JGG, Warrel DA (Eds). Oxford University Press: 3500-3511, 1996 Weaver VM, Each B, Walker PR, Sikorska M: Role of proteolysis in apoptosis- involvement of serine proteases in intemucleosomal DNA fragmentation in immature thymocytes. Biochem. Cell. Biol. 71: 488-500, 1993. Webb SJ, Harrison DJ, WylUe AH: Apoptosis; An overview of the process and its relevance in disease. Adv. Pharmacol. 41: 1-34, 1997 Wellinghausen N, Kirchner H, Rink L: The immunobiology of zinc. Trends Immunol Today. Nov, 1997 White J, Herman A, Pullen AM, Kubo R, Kappler JW, Marrack P: The V beta-specific superantigen staphlococcal enterotoxin B: Stimulation of mature T cells and clonal deletion in neonatal mice. Cell. 56(1): 27-35, 1989 White K, Tahaoglu E, Steller H: Cell killing by the drosoplila gene Reaper. Science. 271: 805-807, 1996 277 Williams CD, Linch CD, Watts MJ, Thomas NS: Characterisation of cell cycle status and E2F complexes in mobilised CD34+ cells before and after cytokine stimulation. Blood 90(1): 194-203, 1997 Williams GT, Smith CA, Spooncer E, Dexter TM, Taylor DR: Haemopoietic colony stimulating factors promote cell survival by surpressing apoptosis. Nature. 343: 76-79, 1990 Wolvetang EJ, Johnson KL, Krauer K, Ralph SJ, Linnane AW: Mitochondrial respiratory chain inhibitors induce apoptosis. FEBS Lett. 339: 40-44, 1994 Wright EG, Lord BI: Haemopoietic tissue. Ballieres. Clin. Haematol 5(3): 499-507, 1992 Wu D, Wallen HD, Nunez G: Interaction and regulation of subcellular localization of Ced-4 by Ced-9. Science. 275: 1126-1129, 1997 Wyllie AH, Kerr JRF, Currie AR: Cell death: the significance of apoptosis. Intern Rev Cytol 68: 251-306, 1980 Wyllie AH, Morris RG, Smith AL, Dunlop D: Chromatin cleavage in apoptosis: Association with condensed chromatin morphology and dependence on macromolecular synthesis. J. Pathol. 142: 67-77, 1984 Xue D, Horvitz HR: Caenorhabditis Elegans Ced-9 protein is a bifunctional cell death inhibitoe. Nature. 390: 305-308, 1997 Yang J, Lui X, Bhalla K, Kim CN, Ibrado AM, Cai J, Peng TI, Jones DP, Wang X: Prevention of apoptosis by Bcl-2: relaese of cytochrome C from mitochondria is blocked. Science. 275(4303): 1129-1132, 1997 Yuan JY, Shaham S, Ledoux S, Ellis HM, Horvitz HR: The C. Elegans cell death gene, Ced-3 encodes a protein similar to the mammalian interleukin 1 p converting enzyme. Cell. 75: 641-652, 1993 Yuan J: Evolutionary conservation of a genetic pathway of PCD. Exp.Pathol. 60(1): 4- 11, 1996 278 Yung J, Lui X, Bhalla K, Kim CN, Ibrado AM, Cai J, Peng-I T, Jones DP, Wang X: Prevention of apoptosis by Bcl-2 release of cytochrome c from mitochondria blocked. Science 275: 1129-1132, 1997 Zalewski PD, Forbes IJ, Giannakis C: Physiological role for zinc in prevention of apoptosis (gene-directed-death). Biochem. Int. 24(6): 1093-1101, 1991 Zalewski PD, Forbes IJ, Betts WH: Correlation of apoptosis with change in intracellular labile Zn(II) using zinquin, A new specific probe for Zn(II). Bichem J. 296:403-408, 1993 Zalewski PD, Forbes IJ, Seamark RF: Flux of intracellular labile zinc during apoptosis (gene directed cell death) by a specific chemical probe, Zinquin. Chem. Biol. 1: 153- 161, 1994 Zaleswski PD: Zinc and immunity: implication for growth, survival and function of lymphoid cells. J. Nutr. Immunol. 4: 39-84, 1996 Zamzami N, Marchetti P, Castedo M, Decaudin D, Macho A, Petit PX, Mignotte B, Kroemer G: Sequential reduction of mitochondrial transmembrane potential and generation of reactive oxygen species in early programmed cell death. J. Exp Med. 182: 367-377, 1995 Zevin S, Link G, Grady RW, Hider RC, Peter HH, Hershko C: Origin and fate of iron mobilised by the 3-hydroxypyridin-4-one oral iron chelators; studies in hypertransfused rats by selective radioprobes of reticuloendothelial and hepatocellular iron stores. Blood 79: 248, 1992 Zhang Y, MarciUat O, Giulivi C: The oxidative inactivation of mitochodrial electron transport chain components and ATPase. J Biol Chem 265: 16330-16336, 1990 Zou H, Henzel W, Lui X, Lutschg A, Wang X: Apaf-1, a human protein homologous to C. elegans Ced-4, participates in cytochrome c dependent activation of caspase-3. Cell. 90: 405-413, 1997 279